Beating burnout: Addressing emotional exhaustion at work

To date, research has approached EE through the lens of resource-based theories such as Stevan Hobfoll’s COR Theory, which defines resources as ‘valued objects, personal characteristics or energies.’ According to COR, employees have a fundamental desire to preserve the resources they have and to gain new ones, as well as to invest their resources to protect from future resource losses. Thus, resource consumption is a defining feature of EE, and the subjective experience of EE is influenced by factors that both consume and provide resources.

In our research, we considered how resource-providing and resource-consuming factors impact not just the level of EE but the form of what we call ‘EE growth curves’ on a daily basis.

Resources differ along two dimensions: the source of the resource as either contextual (outside the self) or personal (inside the self); and the transience of the resource as either volatile (fleeting, dynamic) or structural (durable, ongoing). In our study, we included one resource factor from each of these four quadrants. And because EE is also the result of resource expenditure, we included two resource-consuming predictors.

Following are the factors we looked at as predictors of EE growth curves during a typical workday:

Prior-evening psychological detachment. Psychological detachment from work is defined as ‘an individual’s sense of being away from the workplace’ and is characterized by an absence of work-related thoughts and activities during off-work hours. Detachment has been shown to be a factor that replenishes resources lost due to job-related effort expenditure.

We theorized that prior-evening detachment would serve as a resource-providing experience that aids in the recovery process and therefore should contribute to lower EE at the start of the workday.

We theorized that high detachment the previous evening would produce lower initial levels of EE the next morning followed by an increase during the day. In contrast, days with low detachment in the prior evening would result in elevated initial levels of EE that remain high throughout the day. In essence, we felt that individuals who do not detach in the prior evening would not feel restored at the start of the workday, resulting in higher EE throughout the day.

Socializing with coworkers

During formal or informal breaks, individuals may connect with others, generate positive emotions and build psychological resources. We theorized that the incorporation of these new resources during the workday would slow the natural increase of EE, producing a weaker workday slope.

To the extent that employees have some choice about who they socialize with at work, they are likely to pick higher quality relationship targets, which means socializing should yield more resources than it consumes. Socializing also satisfies one’s need for relatedness, which is associated with greater psychological vitality and engagement, both of which are negatively related to EE.

Supervisor support

COR theory suggests that individuals ‘call on’ the resources they possess to offset the negative effects of resource loss. Supervisor support is one resource that we felt was likely important for preventing and slowing resource loss. Perceived supervisor support reflects employee perceptions that their supervisor values their contributions and is willing and able to provide emotional and instrumental assistance in times of need. Previous research has found that supervisor support is positively associated with psychological meaningfulness and vigor and negatively associated with EE. We felt that when one has a supportive supervisor, it is likely that the overall level of EE experienced at work is lower.

A sense of autonomy

A growing body of evidence highlights the critical role of feeling autonomous versus feeling controlled in one’s actions as important determinants of resource creation and consumption. Behaviours supported by ‘autonomous motivation’ feel self-authored, authentic and invigorating. In contrast, ‘controlled motivation’ is made up of extrinsic motivation, which reflects actions driven by explicit pressure (e.g. punishments or rewards in the form of social and material outcomes.) Behaviours supported primarily by controlled motivation are not aligned with the self and, as a result, consume resources without replenishing.

Interactions with difficult customers

In service roles, it is well-documented that employees must exert emotional energy when interacting with customers and conform to ‘display rules’ to show positive emotions. When customers are rude, employees may experience negative emotions that increase the amount of energy needed to regulate their emotions. We felt that this accelerated expenditure of emotional resources likely contributes to an increase in EE over the workday.

Also read: How can companies rethink the idea of productivity?

Our Research

As indicated, EE is at least partially the result of expending emotional resources in response to demands, which is especially likely in certain occupations. Two defining features of service work are high levels of social interactions (e.g. with customers, clients) and low job control.

In our study, we focused on customer service employees, who face consistently high emotional demands that are likely make EE a daily experience.

We expected that workday EE would start off at relatively low levels and would have a natural tendency to increase throughout the workday as a function of expending resources to meet job demands, and assuming work duties are evenly distributed across the typical workday, that there would be a steady reduction of resources through the day and a corresponding increase in EE.

Call centre participants were recruited in-person during organization-wide ‘town hall’ meetings, and interested employees provided the research team with their e-mail addresses and work schedules. Service industry participants were recruited through social media websites (i.e., LinkedIn, Reddit, Twitter and Facebook) and provided information about their occupation and work schedules via an online form. Eligible participants were sent the initial survey followed by daily surveys.

The employees we recruited worked at least 30 hours per week and all interacted with the public. Participants were asked to complete daily surveys at the start, middle and end of their shift over the course of up to 16 workdays. The three daily surveys were used to model day-level EE growth curves.

Our call centre sample consisted of 64 participants with 502 complete days, and the service industry sample consisted of 50 individuals with 428 complete days. The call centre participants handled inbound and outbound calls for a variety of third-party clients, while the service employees worked in a variety jobs with interactions with the public: 24 per cent in educational services; 22 per cent in healthcare; eight per cent in retail; six per cent in administrative support; six per cent in professional, scientific and technical fields; four per cent in finance; and 30 per cent in a variety of other public-facing occupations. On average, participants were 32 years old, with 79.6 per cent identifying as female.

At the start of shift, middle of shift, and end of shift, participants were asked to rate their agreement with the following items (1 = strongly disagree; 5 =strongly agree): ‘… I feel emotionally drained,’ ‘…I feel used up,’ and ‘I feel burned out.’ Detachment from work in the prior evening was assessed once per day in the daily pre-shift survey. Participants were asked to think about the prior evening and rate their agreement (1 = strongly disagree; 5 = strongly agree) with four items including “…I didn’t think about work at all,” and “… I got a break from the demands of work.”

The extent to which individuals socialized with coworkers was assessed twice per day: once mid-shift and once at the end of the shift. Participants rated the following items on a six-point scale (1 = not at all; 6 = to a very large extent): “I socialized with others” and “I spent time with/talking to friends.”

To determine mistreatment by customers, we used an existing measure that consists of 10 items asking individuals to rate the frequency (0 = never, 4 = all the time) with which they were treated poorly by individuals outside of their organization. Example items were ‘Demanded special treatment’ and ‘Yelled at me. Participants completed the survey once in the mid-shift survey and once in the end-of-shift survey.

Our Findings

Most of the variance in EE was found to be at the between-person-level of analysis, which aligns with the notion that, as a component of burnout, EE reflects the longer-term experience of an individual being emotionally depleted. However, we also found that nearly 30 per cent of the variance in EE was within-the-worker. This suggests that there is a non-trivial amount of within-person variation and that within-day variation in EE relates to how the workday unfolds.

At the day level, we found support for prior-evening detachment, within-day socializing with coworkers and within-day interactions with difficult customers as predictors of within-person variability in EE growth curves. At the worker level, we found support for perceived supervisor support, autonomous motivation for the job and controlled motivation for the job as predictors of EE growth curves.

Our findings have some key implications for leaders—particularly in the service sector, which makes up two-thirds of the Canadian economy. Together, these predictors suggest a variety of potential points of intervention for improving daily EE. Organizations should be aware of the general increase in employee EE throughout the day and how resource-providing and resource-consuming factors can influence this trend, using this knowledge to inform efforts that emphasize activities that replenish resources during the day.

While most of the variance in EE at the start of the day was a function of the individual, most of the variance in how EE changes throughout the day was a function of the day’s events—providing guidance on potential organizational interventions. For example, ‘prior evening detachment’ was found to be a positive predictor of day-level slopes, suggesting that high detachment produces benefits regarding how the workday begins, followed by the expected increase in EE over the workday. In contrast, low detachment the previous night resulted in higher starting levels of EE that remained elevated throughout the workday.

This suggests that low-detachment individuals do not get the break from high EE at the start of the workday that high-detachment individuals receive, resulting in a different pattern of growth over the course of the day. Knowing that detachment benefits may not be sustained throughout the day, organizations should use a variety of interventions in conjunction with encouraging employees to detach in the evenings.

As socialization with coworkers was found to have a buffering effect on increases in EE, organizations may focus on the cultivation of a positive social environment where employees are encouraged to get to know each other, with break opportunities and a comfortable physical location (i.e. break rooms) where positive social interactions can organically occur.

In addition to providing opportunities to socialize, daily work breaks can encourage respite activities aimed at replenishing resources and mitigating the expected increase in EE. Opportunities for non-work-related, enjoyable tasks such as relaxation activities (stretching, meditating) or socializing can aid in resource recovery during the workday.

Because mistreatment by customers exacerbates EE, organizations should consider providing support to help employees cope with difficult interactions. For example, mindfulness or perspective-taking training, with the goal of promoting less emotional reactivity when faced with unpleasant customers. Social support has also been found to moderate the relationship between customer mistreatment and EE, so organizations could focus on efforts to promote communication between employees and their supervisors and coworkers.

Given that supervisor support can mitigate the average increase of EE over a workday, organizations may introduce supervisor training programs, encourage effective and honest communication between employees and supervisors to facilitate climates of trust, and increase leader–member exchange quality to increase perceptions of supervisor support.

Additionally, given the beneficial influence of autonomous motivation for employees’ starting levels of EE, organizations could attempt to increase these motivations by makingchanges to the job and reward contingencies.

To increase autonomous motivation, organizations could encourage employee participation in decision-making processes (where possible), offer choices within structured options, and solicit and acknowledge employee feedback and perspectives. By emphasizing personal choice and voice rather than external control, implementation of such organizational practices can help provide meaningful reasons and increase autonomous motivation for engaging in work tasks.

Finally, understanding how EE may change over the day and in response to various antecedents may inform both individuals and organizations about the optimal timing and scheduling of job demands. For instance, it may be best to schedule difficult demands at the start of the day, when employee EE tends to be low.

In closing

Our research reveals that employee emotional exhaustion usually increases throughout the workday and that these trajectories are influenced by a variety of factors, including previous-night detachment from work, socialization with colleagues and supervisor support.

It is our hope that understanding the variables that relate to these workday growth curves in emotional exhaustion will inform the development of interventions targeting the design of work, social and structural aspects of the environment, and the training of employees on adaptive coping behaviours.

John Trougakos is a Professor of Organizational Behaviour and HR Management in the Department of Management at University of Toronto-Scarborough, with a cross-appointment to the Organizational Behaviour area at the Rotman School of Management. This article summarizes his paper, “Emotional Exhaustion Across the Workday: Person-Level and Day-Level Predictors of Workday Emotional Exhaustion Growth Curves,” co-authored with Faith C. Lee, James M. Diefendorff and Megan T. Nolan. The paper was recently published in the Journal of Applied Psychology.