Intangible rewards and the workplace well-being of civilian Department of National Defence employees

Abstract

Introduction:

Since the onset of the COVID-19 global pandemic, individuals have had an opportunity to re-evaluate the role of work in their lives. Shifts in worker values offer employers the possibility of revisiting their offerings in ways that optimize employee experiences and well-being while also considering employee heterogeneity. These offerings typically include traditional compensation elements, as well as non-financial, or intangible, rewards. Positioning intangible rewards as job resources, the present study explored the moderating influences of employee characteristics on the relationships between these resources and workplace well-being among civilian employees of the Department of National Defence (DND).

Methods:

This study consisted of secondary analyses of the 2022 Public Service Employee Survey microdata, restricted to DND employees. Specifically, latent moderation analyses regressed emotional exhaustion and job satisfaction on employee characteristics (i.e., disability status, leader status, age), intangible rewards (e.g., support for flexible work arrangements [FWAs], a manageable workload), and their interactions.

Results:

Overall, higher levels of recognition were associated with reduced emotional exhaustion, and greater support for FWAs was related to increased job satisfaction. The remaining relationships were more nuanced. For instance, support for FWAs was a stronger protective factor of emotional exhaustion among persons with disabilities relative to persons without disabilities, and workload management was a stronger determinant of job satisfaction among leaders compared with non-leaders.

Discussion:

Notwithstanding its limitations, this study has preliminary implications for research and practice, especially with respect to the recruitment and retention of key workforce segments.

Lay Summary

The COVID-19 global pandemic triggered shifts in worker values, presenting employers with a unique opportunity to revisit their offerings in ways that optimize both employee experiences and workplace well-being. These offerings typically include traditional compensation elements and non-financial, or intangible, rewards. Analyses of the 2022 Public Service Employee Survey examined how three sub-populations of civilian Department of National Defence (DND) employees (i.e., persons with disabilities, leaders, and young employees) experienced intangible rewards (e.g., support for flexible work arrangements, a tolerable workload), as well as the links between their experiences of these rewards and their workplace well-being. The results of this study suggest that, in addition to blanket strategies, targeted workplace interventions may be necessary to improve the workplace well-being of sub-populations of civilian DND employees who differ in their experiences of intangible rewards and in how these experiences affect their workplace well-being.

Introduction

Since the onset of the COVID-19 global pandemic, individuals have had an opportunity to re-evaluate the role of work in their lives and question its meaning, purpose, and congruence with their values.¹ Shifts in worker values, including a greater emphasis on flexibility, organizational support, and well-being, provide employers with the opportunity to revisit their offerings in ways that align with employee expectations.^2,3 How employees experience these offerings is central to their workplace well-being,⁴ as decades of research in the social sciences have underscored.^5–7 In turn, poor workplace well-being is associated with many detrimental outcomes, including costs to employers in the form of lost productivity from absenteeism, presenteeism, and turnover.^8–10

The Federal Public Service Workplace Mental Health Strategy recognizes the importance of workplace well-being in the federal public service (FPS) and reflects a commitment to establishing a healthy, respectful, and supportive workplace.¹¹ To wit, one of its stated objectives is to measure, report, and continuously improve workplace well-being. One of the goals of this study was to support this objective through an exploration of the links between the employer’s offerings and the workplace well-being of civilian employees of the Department of National Defence (DND). Furthermore, this study aimed to highlight heterogeneity in employees’ experiences of these offerings, in their workplace well-being, and in the ways their experiences translated into their workplace well-being.

Total rewards

It is essential for an organization to have an understanding of the reasons why someone should want to become, and remain, an employee of the organization; a concise, clear, and compelling distillation of these institutional advantages is referred to as the organization’s employee value proposition (EVP).^12–14 Reflecting on its EVP will, in turn, require consideration of what exactly an organization’s approach is to total rewards (TR).^12,15 The notion of TR has received considerable attention in the human resources management practice literature in the past two decades.¹⁶

A comprehensive treatment of TR recognizes the need to get financial, or tangible, rewards (e.g., base and overtime pay, health and dental benefits) right but also emphasizes the importance of presenting employees with non-financial, or intangible, rewards, such as gratifying experiences, flexible work arrangements (FWAs), and opportunities to develop their skills and careers.^12,15,16 Organizations can thus optimize their reward expenditures across a spectrum of rewards, to best meet their needs and those of their employees.¹⁶ Although practitioners and employees alike typically consider tangible and intangible rewards holistically, compensation researchers tend to focus solely on either pay or benefits.¹⁷ This disconnect between practice and research has inspired pleas for academic research on TR that moves beyond compensation.^15,18

To these points, TR models from consultancies typically consider a wide assortment of rewards. For instance, in their recent review of more than 20 TR models, Rowley and Blais identified the following categories of rewards in addition to pay and benefits (although these are beyond the scope of the present study):¹⁹ work-life (e.g., flexible and special work arrangements, maternity and parental allowances), training and development (e.g., opportunities for learning, coaching, mentoring, and advancement, career opportunities), recognition and performance (e.g., rewards and recognition, performance management), and work environment (e.g., participation in decision-making, social support, manageable workload). These categories provide a good working model, but they are not definitive. Depending on the organizational context and employees, the implementation of a TR strategy may demand fewer, or more, categories than these.¹⁶

Although it is straightforward for competitors to offer comparable tangible rewards to recruit and retain employees, they cannot as readily imitate intangible rewards.¹² As a result, an employer’s ability to provide an authentic and positive employee experience is becoming a differentiating factor in a market where top talent is in short supply and high demand.¹⁴ This makes intangible rewards particularly powerful in enhancing the value of financial offerings, especially when organizations are under pressure to control or reduce costs.²⁰ For instance, recent studies estimated that employees value working from home at the equivalent of about a 4%−8% wage increase,^21–23 and they are generally willing to forego larger salaries in exchange for more meaningful work.²⁴

Effectively communicating to prospective and current employees the full spectrum, value, and uniqueness of the rewards they receive, including intangible offerings, is a key component of a TR strategy.¹⁶ Often, employees are either unaware of the rewards available to them, or they simply do not think of them as rewards (e.g., training and development, a positive work environment).¹⁶ Clarifying rewards and managing expectations regarding them also helps to ensure that organizations fulfill their obligations to their employees.²⁵

Intangible rewards as job resources

The present study drew from Conservation of Resources (COR) theory and the Job Demands-Resources (JD-R) model.^26–30 According to COR theory, individuals strive to obtain and protect resources — for instance, the “good things” at work^31(p.121) — that can accumulate through a reinvestment process and form surpluses, which in turn improve well-being.^, According to some scholars, when it comes to resource losses and gains, the JD-R model represents an elaborate work-related application of COR theory.^,

Specifically, the JD-R model suggests that job demands deplete employees’ resources, leading to burnout and related outcomes (e.g., ill health, absenteeism, turnover).^28–30 In addition, it posits that job resources stimulate employee motivation, resulting in work engagement and associated outcomes (e.g., superior performance) while simultaneously enabling a more efficient management of job demands.^28–30 Several meta-analyses support these propositions.^36–38 The extant research has also shown associations between job resources and burnout, as well as between job resources and negative outcomes such as sickness absence and intent to turn over.^39–41

Following Conway and colleagues and Li and colleagues,^42,43 the current research positioned intangible rewards as job resources within the JD-R framework. Many studies, including meta-analyses, have highlighted the relationships among employee burnout, work engagement, or their correlates (e.g., job satisfaction, organizational commitment) and the work-life interface,^5,40,44,45 training and development,^5,41,46,47 recognition and performance,⁴⁸ job autonomy or control and participation in decision-making,^{5,37,41,43,49} social support,^5,35,36,50 and a manageable workload or manageable quantitative demands.^5,37,43 In the FPS in particular, recognition and participation in decision-making were, among several job resources, the strongest determinants of job satisfaction across two consecutive iterations of the Public Service Employee Survey (PSES).⁵¹ With respect to the defence team (DT) specifically, the results of the 2018 Defence Workplace Well-Being Survey underscored the importance of perceived organizational support, task significance, and relatedness (i.e., feeling connected to others) to bolstering workplace well-being.^52,53

Employee characteristics

Evidence hints at heterogeneity in employees’ experiences of and preferences for rewards, based on their gender, age, job level, or caregiving status, as well as in the ways these experiences or preferences translate into job choices.^4,26,54 Organizations can address this heterogeneity by tailoring and communicating their offerings in view of their recruitment and retention objectives. For example, to recruit with intention and appeal to top talent, the FPS may want to consider its offerings, how they appeal to different workforce segments, and how they differentiate the FPS from its competitors.⁵⁵ To explore heterogeneity in employee experiences of intangible rewards, in their workplace well-being, and in the ways these experiences translate into workplace well-being (for a similar analysis applied to employees across sectors, see Van den Broeck and colleagues),⁵⁶ the present study focused on three sub-populations of employees, namely, persons with disabilities (PwDs), senior-level public servants, and young employees. Each of these sub-populations currently faces unique challenges in the FPS.

In keeping with historical trends, the representation of PwDs in the core public administration was 6.2% in 2023, whereas their labour market availability was 9.1%.⁵⁷ Despite launching several initiatives, the DND has faced difficulties and encountered barriers to the recruitment of PwDs, especially in some occupations, and, in 2020, the representation of PwDs in the DND was 5.5%.⁵⁸ The DND’s workplace culture and obstacles to career advancement may have contributed to retention challenges among this population.⁵⁸ The extant research has shown that, relative to persons without disabilities, PwDs face less autonomy and flexibility,^59,60 worse relationships with and treatment by management,^61,62 difficulties with colleagues,⁶¹ and poorer workplace well-being.^61–64 Similar findings emerged in the FPS, with employees, including DT members, who identified as PwDs reporting lower levels of various job resources,⁶⁵ as well as lower levels of workplace well-being.^65,66

The Organisation for Economic Co-operation and Development highlighted the need to recruit and retain senior-level public servants (executives in the FPS) with the right skills and competencies, given their key role in achieving any future vision of the public service.⁶⁷ Executives’ workplace well-being influences their behaviours and effectiveness, which, in turn, affects their subordinates’ workplace well-being and associated outcomes,^68,69 making their function all the more important. Although the topic of leader workplace well-being remains largely unexplored,^68–70 the research to date has shown that leaders are not impervious to poor psychological functioning and that there are several reasons why high-quality leadership might come at a high cost to these individuals (e.g., responsibility, social isolation).⁷⁰ In fact, transitioning to a leadership role can be a double-edged sword because it may give rise to both a resource loss (e.g., an increase in time pressure associated with an increase in exhaustion) and a resource gain (e.g., an increase in autonomy related to an increase in job satisfaction).^69,71 In the FPS specifically, the 2021 Executive Work and Health Study showed that emotional exhaustion had reached its highest levels since 2012 — an on-going trend that the pandemic exacerbated.⁷²

Young adults may differ from seasoned workers in their experiences of, or preferences for, intangible rewards. A fuller understanding of how best to recruit and retain employees across the life cycle would likely help employers, including the DND, to do so. For instance, to remain competitive, employers may want to adapt their TR approaches to both the experiences of the mature workers they want to retain and to the preferences of the young workers they need to hire and integrate into their existing workforce.¹⁷

The literature points to an association between age and workplace well-being (in a linear or curvilinear fashion), with older workers experiencing reduced emotional exhaustion and increased job satisfaction compared with young employees.^73–76 Studies also revealed linear or curvilinear associations between age and working conditions, with older workers reporting fewer opportunities to learn new transferable skills but experiencing more co-worker support, a slower pace, more opportunities to work from home, and greater autonomy in work scheduling and content relative to their young counterparts.^22,77,78 As for the moderating effects of age on the relationships between working conditions and workplace well-being, they seem to depend on both the specific working condition and the specific indicator of workplace well-being under consideration.^78,79 This may at least partly explain why the existing literature lacks consensus on the age-specific effects of, for instance, autonomy and flexibility on workplace well-being.^{22, 54, 79,80, 81}

In sum, drawing from the preceding literature, this study addressed the following research questions (RQs):

RQ1: Do intangible rewards, positioned as job resources in the JD-R model, relate to the workplace well-being of civilian DND employees in expected ways?

RQ2: Do disability status, leader status, and age relate to intangible rewards and workplace well-being among civilian DND employees?

RQ3: Do these characteristics moderate the relationships between intangible rewards and workplace well-being among civilian DND employees?

Methods

Participants and procedure

The PSES is a census with a cross-sectional design and voluntary participation, targeting active employees of organizations in the core public administration and participating separate agencies.^82,83 It directly assesses employees’ perceptions of their working conditions and their workplace well-being in their first official language (FOL) of choice. The 2022–23 PSES was available for completion from Nov. 21, 2022, to Feb. 5, 2023.⁸² Spanning 92 departments and agencies, 189,584 employees responded to the 163-item electronic survey for an overall response rate of 53%.⁸⁴ For the DND specifically, 12,681 employees participated in the 2022–23 PSES, for a response rate of 46%.⁸⁵

In late 2022 to early 2023, fewer than one-half of DND employees (42%) self-identified as women, 15% as PwDs, and 12% as racialized (see Table 1). Almost three-quarters (73%) reported English as their FOL, and 57% were between 30 and 54 years of age. Almost all employees (93%) were indeterminate, and 40% worked in the administration and foreign services occupational category. Slightly fewer than one-third had been with the FPS between 3 and 10 years (32%). Last, 43% of employees were working on site at the time of data collection.

Table 1.

Characteristics of the civilian DND employee population

Characteristic	%	SE
Gender
Woman	42	0.30
Other	58	0.30
Person with a disability
Yes	15	0.26
No	85	0.26
Racialized employee
Yes	12	0.25
No	88	0.25
Age group, y
≤24	4	0.18
25–29	6	0.18
30–34	7	0.19
35–39	10	0.20
40–44	12	0.20
45–49	13	0.20
50–54	15	0.21
55–59	17	0.23
≥60	16	0.21
First official language
English	73	0.29
French	27	0.29
Tenure with the federal public service, y
<3	19	0.29
3–10	32	0.31
11–20	30	0.30
>20	19	0.27
Employee status
Indeterminate	93	0.18
Term	4	0.15
Other	3	0.11
Occupational category
Executive	1	0.04
Scientific and professional	17	0.16
Administration and foreign services	40	0.16
Technical	13	0.14
Administrative support	11	0.09
Operational	17	0.22
Other	1	0.08
Work location
On-site	43	0.28
Remote	17	0.23
Hybrid	40	0.30

Note: SE = delta method standard error.

Measures

The PSES microdata included a binary variable representing an employee’s identified disability status, as well as a derived variable capturing their leader status (i.e., executive, middle manager, non-leader). Because the number of executives in the sample was small (n = 144), the executive and middle manager categories were collapsed, resulting in two categories, leader (26%) and non-leader. Similarly, the age band variable was recoded to represent young employees (i.e., age 29 y or younger; 10%), prime-age employees (i.e., aged between 30 and 54 y; 57%), and mature employees (i.e., older than 55 y; 33%).⁸⁶ Respondents rated the survey items on five-point scales ranging from 1 (strongly agree) to 5 (strongly disagree). Their responses were re-coded so that a higher rating or score represents a higher level of the underlying construct. The intangible rewards and job resources were operationalized based on a conceptual and empirical mapping of PSES items:⁵¹

1. Support for FWAs (three items): My immediate supervisor supports the use of flexible work hours (e.g., varying the start and end times of the workday)

2. Support for development (three items): My department or agency does a good job of supporting employee career development

3. Recognition (three items): I receive meaningful recognition for work well done

4. Participation in decision-making (three items): I have opportunities to provide input into decisions that affect my work

5. A manageable workload (five items): I can complete my assigned workload during my regular working hours, and

6. Positive relationships (within one’s work unit; four items): In my work unit, individuals behave in a respectful manner.

Burnout and work engagement are at the core of the JD-R model and represent two states of workplace well-being in this framework.^28–30 In this study, burnout was operationalized with a single-item measure of emotional exhaustion, a component of burnout (i.e., after my workday, I feel emotionally drained). In the absence of a suitable measure of work engagement in the PSES, job satisfaction, one of its correlates,⁴⁷ was operationalized with a single-item measure (i.e., overall, I like my job; for the same operationalization, see Blais and colleagues).⁵¹ Job satisfaction is one of the most important and widely researched variables in organizational psychology, with established links to outcomes such as performance and turnover.⁸⁷

Analytical plan

Stata (version 17.0) was used for the main analyses,⁸⁸ and the weighting factors and 1,000 bootstrap weights provided with the PSES microdata were applied to these analyses.^82,89 The present study used a factor scores approach to latent moderation analysis. Specifically, the factor scores for the intangible rewards were extracted from a preliminary measurement model and saved for use in the main analyses.^90,91 This model was estimated via exploratory structural equation modelling (ESEM)⁹² in Mplus (version 8.10),⁹³ using Blais and colleagues’ model as a starting point.⁵¹ The robust maximum likelihood (ML) estimator was used along with the full information ML algorithm to handle missing responses at the item level.⁹³ Because ESEM in Mplus is not currently available with bootstrap weights, only the weighting factors were applied to these preliminary analyses.

After extracting and saving the factor scores from the preliminary measurement model, the sub-population means were compared, controlling for gender, racialized status, FOL, employee status, tenure with the FPS, and work location. These specific control variables were included in the models on the basis of previous FPS research documenting their associations with the intangible rewards or outcomes in this study.^65,66 Next, full latent moderation models, including all predictors and controls, were tested.^94,95 Specifically, for each moderator, the workplace well-being indicators were regressed on the dummy variables for the moderator, the factor scores for the rewards, and the interaction terms between the dummy variables and the factor scores, controlling for the covariates. A marginal effects approach guided the interpretation of the main and moderating effects.⁹⁴

Results

Preliminary measurement model

The ESEM aligned with Blais and colleagues’ model and resulted in an excellent fit to the data (c298 = 1,621.82, p < .001, root-mean-square error of approximation = 0.035, p = 1.00, 95% confidence interval, 0.034–0.037, comparative fit index = 0.983, Tucker-Lewis Index = 0.964).⁵¹ The standardized target factor loadings ranged between 0.31 (FWAs) and 0.95 (positive relationships; see Table 2), and the six constructs were positively related yet distinct,⁹⁶ with correlations ranging between 0.18 (participation in decision-making and workload management) and 0.72 (recognition and participation in decision-making). The construct reliabilities (i.e., coefficient Hs)⁹⁷ ranged between 0.67 (recognition) and 0.92 (positive relationships).

Table 2.

Standardized factor loadings for the exploratory structural equation modelling solution (unweighted N = 12,676)

Item	SFWAs	SD	R	PDM	WM	PRs
Support for FWAs
Item 1	0.40	0.01	0.28	0.11	0.05	0.14
Item 2	0.31	0.12	0.07	0.12	−0.19	0.22
Item 3	0.78	0.01	−0.03	−0.07	0.16	−0.07
Support for development
Item 1	0.03	0.63	0.11	−0.01	0.07	0.08
Item 2	−0.01	0.88	0.03	−0.19	0.00	−0.08
Item 3	0.03	0.58	−0.13	0.36	0.01	`0.08
Recognition
Item 1	0.09	0.16	0.64	0.02	0.03	0.05
Item 2	0.02	−0.01	0.38	0.28	0.05	0.02
Item 3	0.11	0.06	0.67	0.13	−0.01	0.06
Participation in decision-making
Item 1	0.05	0.12	0.11	0.65	0.00	0.03
Item 2	0.05	0.12	0.05	0.71	0.01	0.04
Item 3	0.02	0.03	0.32	0.43	0.17	0.09
Workload management
Item 1	−0.08	−0.07	0.04	−0.05	0.79	0.05
Item 2	−0.08	0.04	−0.02	0.09	0.64	0.04
Item 3	−0.11	0.12	0.10	−0.03	0.63	0.04
Item 4	0.31	−0.03	−0.11	−0.02	0.59	−0.05
Item 5	0.10	0.02	−0.08	0.09	0.49	−0.02
Positive relationships
Item 1	0.01	0.07	0.06	0.02	0.01	0.71
Item 2	0.00	0.06	0.02	0.03	−0.02	0.78
Item 3	−0.02	−0.07	−0.03	−0.05	0.02	0.95
Item 4	0.11	−0.01	−0.09	−0.03	0.07	0.57

Note: Target factor loadings are in bold, and non-significant factor loadings (i.e., p ≥ 0.05) are in italics.

SFWAs = support for flexible work arrangements; SD = support for development; R = recognition; PDM = participation in decision-making; WM = workload management; PRs = positive relationships.

Mean differences

PwDs reported significantly lower levels of intangible rewards and workplace well-being, relative to persons without disabilities (see Table 3 for the adjusted means; R² = 0.03–0.14). Compared with non-leaders, leaders experienced significantly higher levels of rewards, except for support for FWAs and workload management, on which they fared significantly less positively. Leaders were significantly more emotionally exhausted, yet they were significantly more satisfied with their job. On the one hand, young employees reported significantly greater support for development and a significantly more manageable workload compared with prime-age employees. On the other hand, they reported significantly lower levels of rewards relative to mature employees, except for support for development, on which the two groups did not significantly differ. Compared with prime-age or mature employees, young employees also experienced significantly poorer workplace well-being.

Table 3.

Adjusted means

	Disability status				Leader status				Age group
	Yes		No		Yes		No		Young		Prime		Mature
Variable	b	SE	b	SE	b	SE	b	SE	b	SE	b	SE	b	SE
Intangible reward
Support for FWAs*	−0.14a	0.02	0.06b	0.01	−0.14a	0.01	0.09b	0.01	−0.08a	0.03	−0.04a	0.01	0.18b	0.01
Support for development	−0.15a	0.02	0.09b	0.01	0.17a	0.01	0.01b	0.01	0.10a	0.03	0.01b	0.01	0.11a,c	0.01
Recognition	−0.10a	0.02	0.08b	0.01	0.11a	0.01	0.04b	0.01	0.02a	0.03	0.02a	0.01	0.12b	0.01
Participation in decision-making	−0.09a	0.02	0.07b	0.01	0.19a	0.01	−0.00b	0.01	−0.03a	0.03	0.03a	0.01	0.11b	0.01
Workload management	−0.11a	0.02	0.04b	0.01	−0.37a	0.01	0.16b	0.01	0.02a	0.02	−0.05b	0.01	0.12c	0.01
Positive relationships	−0.15a	0.02	0.09b	0.01	0.10a	0.01	0.04b	0.01	0.05a	0.03	0.02a	0.01	0.12b	0.01
Workplace well-being indicator
Emotional exhaustion†	3.24a	0.03	2.78b	0.01	3.13a	0.02	2.74b	0.01	3.03a	0.04	2.91b	0.01	2.67c	0.01
Job satisfaction‡	4.02a	0.02	4.18b	0.01	4.18a	0.01	4.15b	0.01	3.84a	0.04	4.11b	0.01	4.34c	0.01

Note: b = unstandardized regression coefficient; SE = delta method standard error. Control variables were gender, first official language, tenure with the federal public service, employee status, and work location. Means with different subscripts differ at p < 0.05 (two-tailed).

^*n = 9,981.76 for all the mean comparisons involving the intangible rewards.

^†n = 9,874.92.

^‡n = 9,952.50.

FWAs = flexible work arrangements.

Latent moderation analyses

Disability status

Greater support for development (b = −0.105 [SE 0.015], p < 0.001) and higher levels of recognition (b = −0.088, [SE 0.019], p < 0.001) were each associated with reduced emotional exhaustion overall. The mitigating effect of support for FWAs on emotional exhaustion was stronger (or more beneficial) among PwDs relative to persons without disabilities (see Table 4 for the results of the latent moderation analysis, R² = 0.44–0.48, and Table 5 for the associated marginal effects). In comparison, the mitigating effect of a manageable workload was weaker (or less beneficial) among PwDs. Last, the association between positive relationships and emotional exhaustion was not statistically significant among PwDs. There were no moderating influences of disability status on the various relationships between the intangible rewards and job satisfaction. In general, greater support for FWAs (b = 0.093 [SE 0.012], p < 0.001), higher levels of recognition (b = 0.390 [SE 0.018], p < 0.001), greater participation in decision-making (b = 0.272 [SE 0.015], p < 0.001), more positive relationships (b = 0.075 [SE 0.013], p < 0.001), and a more manageable workload (b = 0.043 [SE 0.011], p < 0.001) were each related to greater job satisfaction.

Table 4.

Predictive results

	Disability status						Leader status						Age group
Variable	Emotional exhaustion*			Job satisfaction†			Emotional exhaustion*			Job satisfaction†			Emotional exhaustion*			Job satisfaction†
	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p
Intercept	2.65	0.35	<0.001	3.93	0.03	<0.001	2.66	0.04	<0.001	3.93	0.03	<0.001	2.64	0.04	<0.001	3.89	0.04	<0.001
Moderator
Mod1	0.29	0.02	<0.001	−0.01	0.02	0.451	0.10	0.02	<0.001	0.00	0.01	0.986	−0.14	0.03	<0.001	0.31	0.03	<0.001
Mod2													−0.23	0.04	<0.001	0.44	0.03	<0.001
Intangible reward
Support for FWAs	−0.18	0.02	<0.001	0.09	0.01	<0.001	−0.24	0.02	<0.001	0.11	0.01	<0.001	−0.37	0.06	<0.001	0.09	0.06	0.114
Support for development	−0.10	0.02	<0.001	0.02	0.02	0.193	−0.12	0.02	<0.001	0.04	0.02	0.036	−0.10	0.07	0.185	0.12	0.07	0.059
Recognition	−0.08	0.02	<0.001	0.38	0.02	<0.001	−0.11	0.02	<0.001	0.40	0.02	<0.001	−0.19	0.09	0.047	0.52	0.08	<0.001
Participation in decision-making	0.01	0.02	0.391	0.28	0.02	<0.001	0.03	0.02	0.082	0.25	0.02	<0.001	0.12	0.07	0.058	0.14	0.06	0.027
Workload management	−0.50	0.01	<0.001	0.05	0.01	<0.001	−0.46	0.02	<0.001	0.03	0.01	0.049	−0.30	0.05	<0.001	0.01	0.05	0.868
Positive relationships	−0.12	0.01	<0.001	0.08	0.01	<0.001	−0.10	0.02	<0.001	0.05	0.01	0.001	−0.16	0.06	0.011	0.18	0.07	0.010
Interaction term
Mod1 × Support for FWAs	−0.17	0.04	<0.001	0.01	0.04	0.885	0.10	0.03	<0.001	−0.05	0.03	0.056	0.19	0.06	0.001	−0.02	0.06	0.785
Mod1 × Support for development	−0.02	0.04	0.685	−0.04	0.04	0.343	0.07	0.03	0.026	−0.08	0.03	0.015	−0.04	0.08	0.640	−0.11	0.07	0.102
Mod1 × Recognition	−0.04	0.05	0.390	0.05	0.05	0.316	0.06	0.04	0.121	−0.02	0.04	0.649	0.11	0.10	0.267	−0.11	0.08	0.172
Mod1 × Participation in decision-making	0.03	0.04	0.477	−0.02	0.04	0.565	−0.06	0.03	0.073	0.07	0.03	0.033	−0.11	0.07	0.092	0.14	0.06	0.027
Mod1 × Workload management	0.14	0.03	<0.001	−0.01	0.03	0.669	−0.06	0.03	0.031	0.06	0.02	0.006	−0.17	0.06	0.002	0.05	0.05	0.343
Mod1 × Positive relationships	0.09	0.03	0.008	−0.02	0.04	0.581	−0.02	0.03	0.568	0.10	0.03	0.001	0.06	0.06	0.356	−0.11	0.07	0.124
Mod2 × Support for FWAs													0.14	0.06	0.017	0.02	0.06	0.804
Mod2 × Support for development													0.04	0.08	0.586	−0.13	0.07	0.051
Mod2 × Recognition													0.12	0.10	0.241	−0.19	0.09	0.030
Mod2 × Participation in decision-making													−0.13	0.07	0.075	0.17	0.07	0.015
Mod2 × Workload management													−0.20	0.06	<0.001	0.02	0.06	0.694
Mod2 × Positive relationships													0.06	0.06	0.352	−0.13	0.07	0.084

Note: b = unstandardized regression coefficient; SE = delta method standard error. Mod1and Mod2 were the prime-age and mature dummy variables, respectively, for the age group analysis, with young as the reference category. Control variables ware gender, first official language, tenure with the federal public service, employee status, and work location.

Mod = moderator, FWAs = flexible work arrangements.

^*n = 9,874.92.

^†n = 9,952.50.

Table 5.

. Marginal effects of the intangible rewards on emotional exhaustion

	Disability status						Leader status						Age group
	Yes			No			Yes			No			Young			Prime age			Mature
Intangible reward	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p
Support for FWAs	−0.35	0.03	<0.001	−0.18	0.02	<0.001	−0.14	0.02	<0.001	−0.24	0.02	<0.001	−0.37	0.06	<0.001	−0.18	0.02	<0.001	−0.23	0.03	<0.001
Support for development	−0.12	0.04	0.002	−0.10	0.02	<0.001	−0.05	0.03	0.050	−0.12	0.02	<0.001	−0.10	0.07	0.185	−0.13	0.02	<0.001	−0.05	0.03	0.030
Recognition	−0.12	0.04	0.005	−0.08	0.02	<0.001	−0.05	0.03	0.081	−0.11	0.02	<0.001	−0.19	0.09	0.047	−0.08	0.02	<0.001	−0.07	0.03	0.025
Participation in decision-making	0.05	0.04	0.265	0.01	0.02	0.391	−0.02	0.02	0.361	0.03	0.02	0.082	0.12	0.07	0.058	0.01	0.02	0.593	0.00	0.03	0.956
Workload management	−0.36	0.03	<0.001	−0.50	0.01	<0.001	−0.52	0.02	<0.001	−0.46	0.02	<0.001	−0.30	0.05	<0.001	−0.48	0.01	<0.001	−0.51	0.02	<0.001
Positive relationships	−0.03	0.03	0.369	−0.12	0.01	<0.001	−0.11	0.02	<0.001	−0.10	0.02	<0.001	−0.16	0.06	0.011	−0.10	0.02	<0.001	−0.09	0.02	<0.001

Note: n = 9.874.92. b = unstandardized regression coefficient; SE = delta method standard error. FWAs = flexible work arrangements. Control variables were gender, first official language, tenure with the federal public service, employee status, and work location.

Leader status

Overall, higher levels of recognition (b = −0.091 [SE 0.019], p < 0.001) and more positive relationships (b = −0.100 [SE 0.013], p < 0.001) were each associated with reduced emotional exhaustion. Among leaders, the mitigating effect of support for FWAs on emotional exhaustion was less beneficial compared with that among non-leaders, and the relationship between support for development and emotional exhaustion did not reach statistical significance (see Table 4 and 5). In comparison, the mitigating effect of a manageable workload on emotional exhaustion was more beneficial among leaders.

Greater support for FWAs (b = 0.092 [SE 0.012], p < 0.001) and higher levels of recognition (b = 0.393 [SE 0.018], p < 0.001) were each associated with higher levels of job satisfaction in general. The enhancing effect of support for development on job satisfaction did not reach statistical significance among leaders relative to non-leaders (see Table 4 for the results of the latent moderation analysis and Table 6 for the associated marginal effects). In contrast, the enhancing effects of participation in decision-making, a manageable workload, and positive relationships on job satisfaction were more pronounced among leaders.

Table 6.

Marginal effects of the intangible rewards on job satisfaction

	Disability status						Leader status						Age group
	Yes			No			Yes			No			Young			Prime age			Mature
Intangible reward	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p	b	SE	p
Support for FWAs	0.10	0.04	0.005	0.09	0.01	<0.001	0.06	0.02	0.013	0.11	0.01	<0.001	0.09	0.06	0.114	0.08	0.01	<0.001	0.11	0.02	<0.001
Support for development	−0.02	0.04	0.635	0.02	0.02	0.193	−0.04	0.03	0.131	0.04	0.02	0.036	0.12	0.07	0.059	0.02	0.02	0.387	−0.01	0.02	0.644
Recognition	0.43	0.05	<0.001	0.38	0.02	<0.001	0.38	0.03	<0.001	0.40	0.02	<0.001	0.52	0.08	<0.001	0.40	0.02	<0.001	0.33	0.03	<0.001
Participation in decision-making	0.25	0.04	<0.001	0.28	0.02	<0.001	0.32	0.03	<0.001	0.25	0.02	<0.001	0.14	0.06	0.027	0.28	0.02	<0.001	0.30	0.03	<0.001
Workload management	0.03	0.03	0.288	0.05	0.01	<0.001	0.09	0.02	<0.001	0.03	0.01	0.049	0.01	0.05	0.868	0.06	0.01	<0.001	0.03	0.02	0.102
Positive relationships	0.06	0.03	0.080	0.08	0.01	<0.001	0.15	0.03	<0.001	0.05	0.01	0.001	0.18	0.07	0.010	0.07	0.01	<0.001	0.05	0.02	0.018

Note: n = 9,952.50. b = unstandardized regression coefficient; SE = delta method standard error. Control variables were gender, first official language, tenure with the federal public service, employee status, and work location.

FWAs = flexible working arrangements.

Age group

Greater support for development (b = −0.104 [SE 0.016], p < 0.001) and higher levels of recognition (b = −0.089 [SE 0.019], p < 0.001) were each associated with reduced emotional exhaustion overall, as were more positive relationships (b = −0.103 [SE 0.013], p < 0.001). Although the mitigating effect of support for FWAs on emotional exhaustion was more beneficial among young employees relative to their prime-age or mature counterparts (see Tables 4 and 5), the mitigating effect of a more manageable workload on emotional exhaustion was less beneficial.

In general, greater support for FWAs (b = 0.091 [SE 0.013], p < 0.001), a more manageable workload (b = 0.045 [SE 0.011], p < 0.001), and more positive relationships (b = 0.074 [SE 0.013], p < 0.001) were each related to higher levels of job satisfaction. On the one hand, the enhancing effect of recognition on job satisfaction was more pronounced among young employees compared with mature employees (see Tables 4 and 6). On the other hand, the enhancing effect of participation in decision-making on job satisfaction was less pronounced among young workers relative to their prime-age or mature counterparts.

Discussion

Summary of the results

The present study supported the Federal Public Service Workplace Mental Health Strategy’s objective of measuring, reporting, and improving workplace well-being. To do so, it drew from COR theory, the JD-R model, and the associated literature to situate intangible rewards as job resources within the JD-R framework and explored the three research questions.

In relation to RQ1 specifically, higher levels of recognition were associated with reduced emotional exhaustion among all employees. This result supports Renger and colleagues’ findings and suggests an avenue for future research,⁹⁸ given the limited evidence connecting recognition at work and emotional exhaustion. Moreover, greater support for FWAs was related to higher levels of job satisfaction among all employees, further solidifying the links between flexibility and positive employee outcomes.^5,44 With respect to the remaining research questions, which tackled employee heterogeneity, the following section covers the results associated with each sub-population in turn. For parsimony, it highlights the moderating influences that were beneficial to each sub-population of employees.

Starting with PwDs, the results align with the literature to date. That is, compared with persons without disabilities, PwDs typically report less favourable working conditions, including reduced flexibility and poorer workplace well-being.^59–66 However, FWAs may help to alleviate their pain and fatigue.⁹⁹ This may at least partly explain why, in this study, support for FWAs was more beneficial in alleviating emotional exhaustion among PwDs relative to persons without disabilities. Future research could investigate the processes underpinning this mitigating influence. Turning to leader status, the results give credence to a dual-pathway model connecting leader status to workplace well-being.⁷⁰ Why leaders experienced more positive relationships compared with non-leaders and why participation in decision-making, a tolerable workload, and positive relationships were more favorable to them relative to non-leaders remain open questions for inquiry, however.

Last, with respect to age, the findings generally support the contentions that young employees experience less flexibility and poorer workplace well-being relative to their older counterparts and that flexibility also plays a stronger role in affecting their workplace well-being.^{22,81,100,101} Generation theory suggests that younger cohorts value employment in which they can flexibly integrate their work and private lives and that, for them, flexibility is not so much a perk anymore as a standard requirement.^100,101 Certainly, given the current debates about flexibility, including hybrid work arrangements, this topic is ripe for further investigation. Young employees also reported lower levels of recognition relative to mature employees, and this intangible reward was a stronger determinant of their job satisfaction relative to either prime-age or mature employees. The extant literature on the associations among age, recognition, and job satisfaction is sparse, and research in this area is needed to better understand these relationships.

Limitations and research implications

First and foremost, future studies would do well to integrate data on pay and benefits with survey data to capture TR more fully and explore how tangible — relative to intangible — rewards affect workplace well-being. Unfortunately, the present study did not do so, because individual-level data on pay and benefits were not readily available for integration with the PSES data. In addition to the intangible rewards in the current study, researchers may want to consider other intangible rewards, such as task significance, which may play a unique role in attracting and retaining FPS employees.¹⁰² A more comprehensive and exciting approach to TR could also leverage mixture modelling to uncover hidden configurations of employee reward experiences and relate membership in these configurations to workplace well-being.¹⁰³

Future research could also consider a wider variety of employee characteristics, such as racialized employee status or Indigenous person status, or intersectionality analyses, which are particularly relevant to the FPS context. Relatedly, it is worth noting that this study did not consider discrimination, which may be central to the experiences of PwDs.⁶¹ In fact, subtle forms of discrimination (e.g., micro-aggressions, condescension) may contribute to retention challenges at the DND specifically,⁵⁸ something future research could investigate. Moreover, future work could study middle managers and executives separately, because they occupy different positions in organizational hierarchies, with middle managers taking on an ambiguous position as a buffer between executives and staff.¹⁰⁴ These different positions, along with distinct compensation schemes, may in turn result in divergent workplace experiences and well-being.

Although the results provide additional support for the validity of the measures in the present study, their psychometric properties remain uncertain, as Blais and colleagues had previously noted.⁵¹ Moreover, some of the constructs (e.g., recognition, FWAs) showed less-than-desirable reliabilities. Together, these caveats suggest that future research would benefit from relying on more established single- or multi-item measures. Last, in the absence of a measure of work engagement in the PSES, this study relied on job satisfaction, a well-known correlate of work engagement, to reflect workplace well-being. Work engagement, along with burnout, remains at the core of the JD-R model, and future studies would do well to consider this state of workplace well-being.

This study relied solely on self-report data, a limitation researchers could address through the integration of informant ratings of workplace well-being (e.g., colleagues, supervisors, spouse) and objective workplace well-being outcomes (e.g., absenteeism, presenteeism, turnover). Furthermore, its cross-sectional design makes it impossible to confirm the directionality of the associations among intangible rewards, moderators, and workplace well-being. It would be fruitful to consider longitudinal designs in future research. A cross-sectional design cannot differentiate between age-related influences (e.g., life or career stages, physical maturation) and generational effects,¹⁰⁵ so the findings should not be interpreted as such. Last, the results of this study apply to the civilian employees of the DND, an organization in which these employees encounter unique working conditions and challenges.¹⁰⁶ Future research could investigate whether the results generalize to other FPS organizations, as well as to other organizational settings, nationally and internationally.

Practical implications

The results suggest that the DND is unlikely to be successful in improving the workplace well-being of its civilian employees through one-size-fits-all workplace interventions. Instead, the organization may want to consider both blanket interventions (e.g., strengthening unit-level relationships) and interventions tailored to its civilian employees’ characteristics and experiences of rewards. For instance, although the increased provision of non-monetary forms of recognition (e.g., a simple, heartfelt thank you from one’s team members, the delivery of regular, accurate, and specific feedback on task performance from one’s leader)^107,108 may be beneficial to all civilian DND employees, young employees may gain more from such practices than mature workers.

Young employees, along with PwDs, may benefit the most from increased support for FWAs, and leaders, from adjustments to their workload. Although accommodations and FWAs are already available to civilian DND employees, some FWAs may not be popular, or the organizational culture may hinder their use.^109,110 For example, when leaders appear unsupportive of flexibility, employees may worry about negative consequences.¹⁰⁹ To alleviate these concerns, leaders may consider engaging in work-life supportive behaviours, such as role modelling and providing employees with greater autonomy in the performance of their functions.¹¹¹

Last, adopting a right-to-disconnect policy may be associated with a more tolerable workload and in turn improved workplace well-being.¹¹² The provision of additional resources to ease demands on employees in times of high pressures and the streamlining of work may also help to mitigate a high workload.^113,114 For instance, training leaders to critique inefficient and time-consuming practices and identify new ways of working more efficiently (e.g., reducing time spent in meetings) to achieve the same objectives may positively affect their workplace well-being.¹¹⁴

The findings also have implications for recruitment. In an era when workers are increasingly mobile and talent is scarce, especially in key workforce segments, organizations ought to shape their EVP to the individuals they want to attract and recruit.¹⁴ More important, they need to offer an aligned employee experience to ensure the promised EVP has lived up to expectations.¹⁴ For example, to attract young workers, the DND’s recruitment efforts may want to emphasize the opportunities for development that will be accessible to these workers after they join the department.

Biography

Ann-Renée Blais, PhD, MA, BSc, received her PhD in quantitative psychology with a specialization in economic and behavioural decision analysis from The Ohio State University. She also completed post-doctoral work in behavioural economics and, more recently, a graduate diploma in public policy and program evaluation. She first worked at the Department of National Defence, and after studying workplace well-being for several years, she established Statistics Canada’s Organizational Health Team in 2020. She is now a Senior Advisor at the Treasury Board of Canada Secretariat where she pursues her interests in statistical modelling and organizational psychology, with a focus on the recruitment and retention of in-demand talent.

Funding Statement

No funding was received for this article.