Emotional Communication and Human Sustainability in Professional Service Firms (PSFs)

Abstract

This study examines the role of work-related emotional communication in promoting the well-being and sustainability of professionals working for professional service firms (PSFs), which depend upon the well-being of their professionals for their own organizational sustainability. Using survey data from 1465 attorneys, a structural equation model was tested including key work-related emotional communication variables as mediators between a dichotomous variable of professional seniority and three dimensions of burnout. Results showed that more experienced attorneys’ reliance on automatic regulation over surface acting has a significant effect on reported burnout. There is no difference based on professional seniority in use of deep acting or communicative responsiveness. We conclude with a discussion of the findings and theoretical and practical implications, as well as provide suggestions for future research.

1. Introduction and Literature Review

Organizational sustainability is about more than environmental impact and economic development. In addition to its economic and ecological resources, an “organization sustains its functional capacity by promoting the sustainability of its human and social resources” [1] (p. 234). Human resources could be described as the individual-level skills and commitments of the organization’s members. Social resources are how those individual human resources are wedded to create networks, communication flow, and organizational climate and culture. Recognizing the social dimensions of sustainability, the United Nation’s World Commission on Environment and Development (also known as “Brundtland Commission”) emphasized providing for the needs of current generations without jeopardizing the ability of “future generations to meet their own needs” [2] (p. 53), quoting [3] (p. 43). These social dimensions of sustainability are also at the heart of organizations’ prioritization of corporate social responsibility (CSR) and stakeholder relationships [4,5], including concern for the long-term well-being of their organizational members (i.e., employees or workers) [6,7].

Overall, the “necessity of a healthy and motivated workforce” requires organizations to actively promote workplace health and safety [5] (p. 812). With greater awareness of their importance in the past decade, many organizations have introduced new programs and committed additional resources to wellness initiatives [8]. In doing so, organizations benefit by better ensuring a more compliant workforce, employee retention, increased productivity, and improved innovation, profits, and growth [4,6]. This investment can pay off in sustainability benefiting long-term organizational outlooks and employees.

The relevance of human sustainability and its relationship to organizational sustainability is especially evident in the context of professional service firms (PSFs; e.g., accounting, law, and architecture firms). Humans are the primary resource in PSFs [9] because their bottom line “depends on the knowledge and competence of human agents” [1] (p. 53). Therefore, nurturing and maintaining human capital is essential for a PSF’s future. For example, Massaro et al. [9] found that human sustainability practices were positively related to accounting firms’ abilities to innovate because they fostered creativity and intellectual capital among their accountants, resulting in a positive effect on overall firm performance.

A prevalent threat to the well-being and sustainable performance of PSF professionals is stress and burnout [2,5,10]. Burnout consists of one or more of three dimensions—exhaustion, cynicism, and inefficacy—which are outcomes of chronic exposure to stressors at work [11]. The rates of burnout can be especially high in professional service occupations because of intense workloads, inflexible deadlines, and interactions with coworkers, clients, and other stakeholders that are often emotionally charged, involving emotional labor or emotional work [12]. Emotional labor is the regulation or fabrication of emotions so that work-related displays of emotion comply with the expectations of one’s employment [13]. It includes three types of emotional communication—surface acting (i.e., faking the expected emotion display), deep acting (i.e., internalizing the appropriate emotion to match the expected emotion display), and automatic regulation (i.e., unconsciously regulating genuinely felt emotion to the expected magnitude of its expression). Comparatively, emotional work [14] is work that engenders genuinely felt emotions and includes the need to communicate compassion or empathy. The communicative construct that is at issue in emotional work is communicative responsiveness (i.e., verbal and nonverbal communication that meets the empathic needs of another person). Each of these communicative acts have been linked to worker burnout.

Studies that examine profession-specific causes of burnout remain rare [10,15,16]; however, framed by Job Demand-Resources (JD-R) Theory [15,17], Powers and Myers [18] recently investigated the influence of work-related emotional communication on attorney burnout. Of the three types of emotional labor (surface acting, deep acting, and automatic regulation), they found that only surface acting increased burnout (via direct effects on the exhaustion and cynicism dimensions of burnout), while automatic regulation decreased the inefficacy dimension of burnout. They also found that the empathic communication required in emotional work to satisfy the expectations of others (communicative responsiveness) reduced the inefficacy dimension of burnout. Pursuant to the JD-R theory, Powers and Myers [18] found some forms of work-related emotional communication function as job demands (stressors) that cause individual strains that, although mitigated by various job resources, eventually lead to burnout. For example, attorneys who feel they must continuously display emotions that they do not actually feel experienced more exhaustion in their work. Finally, Powers and Myers [18] found no significant influence of deep acting on any of the dimensions of burnout.

In designing the current study, the authors speculated that the impact of some job demands may be reduced over time as more senior professionals learn to manage resources and stress more effectively. This speculation is based upon Brewer and Shapard’s [19] meta-analysis findings of a small negative correlation between years worked in a field and the exhaustion dimension of burnout, as well as studies showing that more seasoned attorneys (based primarily on age) may experience less job stress [20,21]. In the latter cases, age likely was a proxy for the attorneys’ professional seniority (i.e., number of years working in the profession)—a variable that better captures career-related experience than a professional’s age [22]. This, coupled with the high attrition rate of attorneys in private practice [23], suggests that attorneys who achieve greater professional seniority may include a subset of attorneys who remained in the practice of law because they are better equipped to cope with exposure to job stressors prevalent in the practice of law, leading to less strain (e.g., burnout). Thus, in its examination of work-related emotional communication and attorney burnout, the current study focused on the total number of reported practice years as a means of assessing professional seniority in the practice of law (highly correlated with age, nonetheless).

More specifically, the current study investigated the roles of various forms of work-related emotional communication on human sustainability in the practice of law, asking: Do attorneys who achieve seniority in the legal profession rely upon different forms of work-related emotional communication than do junior attorneys, therefore leading to less burnout? To answer this question, we focused on two groups of attorneys—junior attorneys who have practiced 10 years or less, and senior attorneys who have practiced 23 years or more—to assess the influence of professional seniority on the three dimensions of burnout, with the following work-related emotional communication variables as mediators: communicative responsiveness, surface acting, deep acting, and automatic regulation (Hypothesized Model; see Figure 1).

Because of the aforementioned evidence that more senior attorneys may experience less job stress and the negative correlation found between years worked in a field and the exhaustion dimension of burnout [19], we predicted that the more senior group of attorneys would experience less burnout via a direct negative influence on exhaustion (H1a). Additionally, given that substantive knowledge of the law as it is practiced is largely learned after law school through work experience as a lawyer [24], we predicted a negative relationship between professional seniority and inefficacy (H1b). Law schools teach law students to recognize legal issues, research the law, and develop legal analysis through the application of existing law to a set of facts. Law students are not taught the contextual and interpersonal demands of the actual practice of law [24]. Wald [24] emphasized the importance of context in legal practice, such as the type of law (e.g., criminal versus civil) and the type of client (e.g., individual versus corporation), stating as an example:

It matters that clients in dissolution-of-marriage cases tend to be emotionally vulnerable and angry, and they often can benefit from legal advice that provides moderation, reasonableness, and a long-term perspective, along the lines of ‘your soon to be ex is going to be your co-parent for many years to come, and while you may have good reasons to be hurt and angry, you may wish nonetheless to be reasonable’ (p. 317).

Given that neither law students nor lawyers receive formal training on emotion in the practice of law [25], it is only through work experience that lawyers learn strategies and tactics to best manage the demands of emotional communication and regulation. Thus, we predicted that senior attorneys would be more adept at conveying the emotion display needs of their practice, whether by internalizing the necessary emotion for the more authentic display of emotion that is characteristic of deep acting (H2a), responding with appropriate expressions of empathy as required for communicative responsiveness (H2b), or tempering their own emotion via automatic regulation (H2c). Therefore, with lawyers predicted to become more adept at deep acting, communicative responsiveness, and automatic regulation through their legal practice experience, we also predicted senior attorneys would have less need to rely on surface acting in their practice (H2d).

The paths between the emotional communication variables and burnout dimensions in the Hypothesized Model were predicted based, in part, on the findings of Powers and Myers [18], as described above. Accordingly, in the current study, the following paths were predicted: (H3a) automatic regulation negatively influencing inefficacy, (H3b) surface acting positively influencing exhaustion and (H3c) cynicism, and (H3d) communicative responsiveness negatively influencing inefficacy.

Although Powers and Myers [18] found no significant effect of deep acting on any of the dimensions of burnout, we proposed to test our less complex model, with more statistical power, with predictions based on the broader literature. Accordingly, per the resource depletion rationale [26], we predicted that (H4a) deep acting would positively influence exhaustion due to the cognitive effort it takes to internalize the emotion that matches the expected emotion display, and that effectively doing so would result in (H4b) deep acting negatively influencing inefficacy. The latter prediction is based not only upon a logical inference that effectively meeting the emotion display expectations of one’s job should reduce feelings of inefficacy, but also because individuals on the receiving end of emotional communication are known to react more favorably to the authentic emotion displays of deep acting [26,27].

Therefore, based on the foregoing, we predicted the following indirect effects: attorneys with greater professional seniority experience less burnout than more junior attorneys due to their greater use of (H5a) automatic regulation, (H5b) deep acting, and (H5c) communicative responsiveness, with each leading to reduced inefficacy. Additionally, more senior attorneys experience less (H5d) cynicism and (H5e) exhaustion by virtue of their lesser reliance on surface acting. However, more senior attorneys may experience (H5f) increased exhaustion by their use of deep acting.

In the following sections, we first describe the research and statistical methods employed to test our Hypothesized Model. We then report on the findings therefrom, which generally support and expand upon the existing scholarship described above. In brief, our findings show differences in work-related emotional communication based on professional seniority that contribute to greater reported burnout among junior attorneys than more seasoned attorneys. After presenting our findings in more detail below, we address the theoretical and practical implications of this study’s findings, as well as recommend additional research avenues that may validate our results in broader and more diverse samples and overcome some of this study’s limitations.

2. Method

2.1. Procedures

Participants were recruited via emails sent to the members of a southeastern U.S. state bar association (n = 81,131). Each email briefly described this study’s purpose, provided a link to the questionnaire administered via Qualtrics, an online survey administrator, and enumerated eligibility requirements. To participate in the survey, members had to be currently (1) practicing law; and (2) representing or counseling clients. A total of 2605 members completed the survey. Of that total, 250 were disqualified for not meeting the foregoing eligibility requirements and an additional 14 were eliminated due to spurious response patterns and data cleaning processes (see the Analytic Approach section below), resulting in 2067 eligible survey takers.

2.2. Participants

The current study focuses on the relationships between work-related emotional communication and attorney sustainability. To isolate the effects of more or fewer practice years (i.e., professional seniority) on work-related emotional communication and, ultimately, burnout, we isolated two participant groups by calculating the mean (M = 17.30) and standard deviation (SD = 12.72) of practice years reported by the original sample of 2067 attorneys. The practice year cutoffs of 10 years or less (Group 1, n = 818) and 23 years or more (Group 2, n = 647) were based on one-half standard deviation from either side of the mean. Accordingly, the final sample size for this study is 1465.

Of the total number of participants, 82.3% were working between 35 and 65 h on average per week, with 11.3% working less than 35 h and 6.4% working more than 65 h. The majority (85.1%) self-identified as White, with the remainder identifying as Native American (0.8%), Asian (2.0%), Black (4.2%), Hispanic (10.4%), Pacific Islander (0.3%), and other (2.4%). Seven hundred and eighty-six (53.7%) of the participants self-identified as male and 674 (46%) self-identified as female.

Comparing the two groups on average hours worked per week, 6% of Group 1’s participants reported working fewer than 35 h per week, while 18% of Group 2’s participants reported the same. Conversely, 38% of Group 1’s participants reported working between 45 and 54 h per week, while only 28% of Group 2’s participants reported the same. Similar percentages of participants within groups reported working between 35 and 44 h per week (~30%), and more than 65 h per week (~6%). As to Group gender distributions, Group 1 consisted of 326 self-identified males (40%) and 489 self-identified females (60%). Comparatively, Group 2 had an opposing gender distribution pattern, comprising 460 self-identified males (71%) and 185 self-identified females (29%). Finally, 92% of Group 1’s participants were 44 years and younger, while 100% of Group 2’s participants were 45 years and older.

2.3. Measures

In addition to this study’s focal variables separately described below (see also

Table 1. Measures Used, Prior Cronbach’s Alpha Coefficients, and Example Items.

Variable	Subscale	Example Item
Emotional Communication	Surface Acting (3 items, α = 0.85, [27]) 1	“In my communication with clients/others on behalf of clients, I…Resist expressing my true feelings.”
	Deep Acting (3 items, α = 0.82, [27]) 1	“In my communication with clients/others on behalf of clients, I…Really try to feel the emotions I have to show as a part of my job.”
	Automatic Regulation (4 items, α = 0.72, [34]) 1	“In my communication with clients/others on behalf of clients, my emotions…Spontaneously coincide with what the interaction with the person(s) calls for.”
	Communicative Responsiveness (6 items, [12]; α = 0.83, as adapted by [35,36]) 2	“I usually respond appropriately to the feelings and emotions of others.”
Burnout	Exhaustion (5 of 16 items; α = 0.84–0.90, [37])	“I feel tired when I get up in the morning and have to face another day of work.”
	Cynicism (5 of 16 items; α = 0.74–0.84, [37])	“I just want to do my job and not be bothered.”
	Professional Efficacy (reverse coded) (6 of 16 items; α = 0.70–0.78, [37])	“In my work, I feel confident that I’m effective at getting things done.”

Note.¹ All scale items were modified to emphasize communication with clients or with others on behalf of clients rather than with patients; ² the item “I show feelings to clients that are different from what I feel inside,” was modified to read: “I show empathy or compassion to others even when they are different from what I feel inside,” to reduce the likelihood that the item would cross-load onto surface acting.

), a host of other demographic and work-related variables were assessed. To assess demographics, participants were asked to report their age, gender, race/ethnicity, average hours worked per week, and number of years of active practice. Work-related variables included constructs that have been consistently linked with burnout across studies [28] and include workload (5 items) [29], role conflict (8 items) [30], job control (1 item) [31], and supervisor, coworker, and familial support (4 items each) [32]. Of the foregoing variables, only two were included as controls: coworker and familial support, both of which are not expected to fluctuate due to the relative seniority of the groups in the practice of law. The others were not controlled for in this study’s analyses because, as explained by Spector et al. [33], doing so would likely remove the variable effects of interest—namely, how professional seniority (i.e., junior versus more seasoned attorneys) might differentially relate to burnout and potentially explain attorney sustainability (see also [22]).

In explanation, gender, age, and the remaining work-related variables are inherently captured by the seniority groups. For example, the more senior group consists of older and predominately male participants as compared to the less senior group. Furthermore, workload, role conflict, and supervisory support are likely to lessen as one advances through their legal career, while job control is likely to increase. Thus, controlling for their effects would remove variance in burnout attributed to potential group differences [22,33].

2.4. Emotional Communication Variables

To measure the extent that each participant engaged in surface acting and deep acting, a 6-item emotional labor scale developed by Brotheridge and Lee [27] was used. Each subscale consisted of 3 items and 5 response options ranging from (0) never to (4) always. Prior research has supported the validity and reliability of both subscales [18,27]. To assess the degree that participants engaged in automatic regulation was measured by a 4-item scale that was developed and validated by Martínez-Iñigo et al. [34] (see also [18]). Finally, communicative responsiveness was measured by a 6-item scale originally developed by Miller et al. [12]. This study, however, used an adapted version of the Miller scale that is psychometrically sound (see Miller and Koesten [35]).

2.5. Burnout Dimensions

To measure the three dimensions of burnout, the Maslach Burnout Inventory–General Survey (MBI–GS) [36] was used, which consists of 16 items total. Both the exhaustion and cynicism inventories have 5 items each, whereas the professional efficacy (reverse coded) inventory consists of 6 items. Previous studies have provided evidence that each of the three dimensions, as measured by the MBI-GS, is valid and reliable (e.g., [18,37]). Data collected on all three burnout dimensions reflect a 7-point Likert-type response scale ranging from (1) strongly disagree to (7) strongly agree.

2.6. Analytic Approach

Using IBM SPSS Statistics version 22 (Armonk, NY, USA) the data were inspected to correct for errors, identify and replace missing values, and test the parametric assumptions required for the statistical analyses addressed below. To test our Hypothesized Model (Figure 1), we used the robust maximum likelihood (MLR) estimator in Mplus 8.3 [38], which proceeded in two steps. First, we assessed the validity of the Measurement Model via a confirmatory factor analysis (CFA), whereby each item was loaded onto its factor. The Measurement Model consisted of two latent control variables (i.e., coworker and familial support), four latent predictor variables (i.e., communicative responsiveness, surface acting, deep acting, and automatic regulation), and three latent outcome variables (i.e., exhaustion, cynicism, and inefficacy). Next, we evaluated the Hypothesized Model (i.e., structural equation model), which consisted of the foregoing nine latent variables, plus one observed variable (i.e., a dichotomous professional seniority group variable that was based on the participants’ reported years of practice).

Measurement and structural equation models are considered a good fit to the data if the chi-square statistic is not significant (p > 0.05), indicating that the Hypothesized Model and a perfectly fitting model are not significantly different from each other. The chi-square statistic, however, is quite sensitive to sample size. That is, in large samples even slight discrepancies between the two models can produce significant chi-square statistics (p < 0.05). Given this study’s large sample size, we examined alternative indices of model fit, including the Root Mean Square Error of Approximation (RMSEA) and its 90% confidence interval, the Standardized Root Mean Square Residual (SRMR), the Comparative Fit Index (CFI), and the Tucker–Lewis Index (TLI). RMSEA and SRMR values of 0.06 and lower are indicative of good model fit to the data [39], although some argue that values less than 0.10 are adequate (e.g., [40]). CFI and TLI values of 0.95 and higher indicate good model fit to the data, while somewhat lower values can suggest adequate fit (≥0.90; [41]). Finally, chi-square difference tests were used to compare the relative fit to the data of the Hypothesized Model to competing, alternative models.

3. Results

Descriptive and inter-item reliability statistics (Cronbach’s alpha) for and correlations among this study’s focal variables are set forth in

Table 2. Descriptive and Inter-item Reliability Statistics for and Correlations among this Study’s Focal Variables.

Variables	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
1. Age (range)	--	--	--
2. Gender	--	--	−0.32 **	--
3. Average Hours per Week (range)	--	--	−0.16 **	−0.05	--
4. Practice Years	17.56	14.92	0.92 **	−0.36 **	−0.15 **	--
5. Seniority Groups	--	--	0.87 **	−0.31 **	−0.13 **	0.94 **	--
6. Job Control	2.42	0.71	0.39 **	−0.14 **	−0.17 **	0.39 **	0.37 **	--
7. Workload	3.14	1.01	−0.21 **	0.07 **	0.50 **	−0.18 **	−0.17 **	−0.23 **	(0.87)
8. Role Conflict	2.89	1.34	−0.22 **	0.09 **	0.17 **	−0.23 **	−0.22 **	−0.38 **	0.35 **	(0.86)
9. Supervisor Support	1.88	1.61	−0.48 **	0.11 **	0.15 **	−0.46 **	−0.44 **	−0.38 **	0.11 **	0.10 **	(0.97)
10. Coworker Support	2.91	1.19	−0.13 **	−0.01	0.17 **	−0.08 **	−0.06 *	−0.08 **	0.14 **	−0.08 **	0.38 **	(0.94)
11. Familial Support	3.34	0.81	−0.13 **	0.05	−0.01	−0.10 **	−0.11 **	0.03	−0.04	−0.09 **	0.17 **	0.28 **	(0.85)
12. Surface Acting	1.78	0.76	−0.17 **	0.08 **	0.07 **	−0.16 **	−0.14 **	−0.23 **	0.16 **	0.32 **	0.04	−0.07 *	−0.08 **	(0.76)
13. Deep Acting	1.65	1.03	0.00	0.02	0.02	−0.01	−0.04	0.04	0.07 **	0.04	−0.00	0.01	0.08 **	−0.00	(0.90)
14. Automatic Regulation	2.74	0.63	0.13 **	0.01	−0.07 **	0.11 **	0.11 **	0.22 **	−0.08 **	−0.24 **	−0.07 *	0.06 *	0.10 **	−0.28 **	0.09 **	(0.82)
15. Communicative Responsiveness	2.77	0.57	0.06 *	0.15 **	−0.00	0.04	0.05 *	0.12 **	0.00	−0.12 **	−0.03	0.06 *	0.09 **	−0.14 **	0.14 **	0.33 **	(0.67)
16. Exhaustion	3.98	1.64	−0.29 **	0.18 **	0.23 **	−0.29 **	−0.27 **	−0.37 **	0.44 **	0.46 **	0.05	−0.11 **	−0.19 **	0.32 **	0.02	−0.28 **	−0.07 **	(0.92)
17. Cynicism	3.46	1.57	−0.18 **	0.06 *	−0.03	−0.17 **	−0.16 **	−0.32 **	0.11 **	0.43 **	−0.01	−0.18 **	−0.17 **	0.34 **	−0.03	−0.31 **	−0.17 **	0.64 **	(0.87)
18. Inefficacy	2.08	0.80	−0.22 **	0.04	−0.01	−0.24 **	−0.23 **	−0.35 **	0.01	0.28 **	0.09 **	−0.12 **	−0.12 **	0.25 **	−0.05 *	−0.38 **	−0.26 **	0.34 **	0.49 **	(0.81)

Note. ** p < 0.01 level (2-tailed); * p < 0.05 (2-tailed); inter-item reliability statistics along the diagonal; Gender: 1 = male, 2 = female; Seniority Groups: 1 = ≤10 practice years, 2 = ≥23 practice years.

.

3.1. Model Estimation

The CFA indicated that all items loaded significantly on their respective factors. However, the hypothesized Measurement Model, which included nine latent variables, was not a great fit to the data (see

Table 3. Measurement and Hypothesized Model Comparisons and Fit Statistics.

Model Comparisons	$\mathit{\chi }$2	df	CFI	TLI	RMSEA, [90% CI]	SRMR
Measurement Model	4290.574 *	629	0.89	0.87	0.063 [0.061, 0.065]	0.048
Measurement Model + Modification Indices	2415.718 *	621	0.95	0.94	0.044 [0.043, 0.046]	0.044
Hypothesized Model	3186.106 *	673	0.92	0.92	0.050 [0.049, 0.052]	0.085
Alternative Model a	3018.295 *	669	0.93	0.92	0.049 [0.047, 0.051]	0.067
Final Model b	2971.873 *	668	0.93	0.92	0.049 [0.047, 0.050]	0.065

Note. * Chi-square statistic ($\chi$²) is significant at p < 0.01; ^a Hypothesized Model with additional model parameters suggested by modification indices; the Alternative Model is a significantly better fit to the data than was the Hypothesized Model (∆$\chi$² (4) = 167.81, p < 0.001); ^b Alternative Model minus insignificant model parameters; the Final Model is a significantly better fit to the data than was the Alternative Model (∆$\chi$² (1) = 46.42, p < 0.001).

). A review of the modification indices suggested that adding residual covariances between items within the same measure would significantly improve model fit. A careful review of the items’ content confirmed that each pair of items (see

Table 4. Residual Item Covariances added to this Study’s Final Measurement Model.

Residual Covariances	M.I.	Item Stem
Exhaustion Item 1 &		I feel emotionally drained from my work.
Exhaustion Item 2	447.90	I feel used up at the end of the workday
Cynicism Item 1 &		I have become less interested in my work since I started this job.
Cynicism Item 2	1150.16	I have become less enthusiastic about my work.
Cynicism Item 2 &		I have become less enthusiastic about my work.
Cynicism Item 5	279.11	I doubt the significance of my work.
Cynicism Item 4 &		I have become more cynical about whether my work contributes to anything.
Cynicism Item 5	693.52	I doubt the significance of my work.
Cynicism Item 1 &		I have become less interested in my work since I started this job.
Cynicism Item 5	206.30	I doubt the significance of my work.
Cynicism Item 1 &		I have become less interested in my work since I started this job.
Cynicism Item 4	198.44	I have become more cynical about whether my work contributes to anything.
Cynicism Item 2 &		I have become less enthusiastic about my work.
Cynicism Item 4	180.60	I have become more cynical about whether my work contributes to anything.

Note. M.I. = Modification indices, which indicate the decrease in the chi-square value if included into the model being tested.

) was similar in content and generally asked the same question in a different way (i.e., reflexive). Thus, it made theoretical sense to covary their error terms [42]. The revised Measurement Model was a good fit to the data (see Table 3).

To test the Hypothesized Model (see Figure 1), we added the observed variable to the model (i.e., the professional seniority group variable) as an additional predictor variable. Overall, the Hypothesized Model was not a good fit to the data (see Table 3). A review of the modification indices suggested adding four additional parameters to the Hypothesized Model, specifically to regress (1) exhaustion on automatic regulation; (2) cynicism on communicative responsiveness, (3) cynicism on automatic regulation; and (4) inefficacy on surface acting (Alternative Model). We tested the Alternative Model and found it to be a significantly better fit to the data than the Hypothesized Model (see Table 3). For parsimony purposes, non-significant parameters were removed from the model, which also significantly improved the model’s fit to the data (Final Model; $\chi$² = 2971.873 (668), p < 0.05); CFI = 0.93; TLI = 0.92; RMSEA = 0.049; SRMR = 0.065; also see Table 3).

3.2. Hypothesis Testing

Overall, the Final Model supports our overarching hypothesis that seniority in the practice of law influences dimensions of burnout both directly and indirectly via emotional communication (See Figure 2). To clarify, participants were divided into two groups based on professional seniority: (a) Group 1: those who have practiced 10 years or less (coded as 1); and (b) Group 2: those who have practiced for 23 years or more (coded as 2). In full support of H1a and H1b, more senior attorneys reported lower levels of exhaustion (β = −0.20, p < 0.01) and inefficacy (β = −0.18, p < 0.01) than more junior attorneys. More senior attorneys additionally reported higher levels of automatic regulation (β = 0.13, p < 0.01) and lower levels of surface acting (β = −0.12, p < 0.01) than reported by more junior attorneys, in support of H2c and H2d, respectively. Contrary to H2a and H2b, seniority in the practice of law was not significantly related to deep acting or communicative responsiveness.

Next, we tested the effects of the four emotional communication variables on the three dimensions of burnout. First, automatic regulation was negatively related to inefficacy (β = −0.36, p < 0.01), in full support of H3a. Second, surface acting was positively related to exhaustion (β = 0.26, p < 0.01) and cynicism (β = 0.30, p < 0.01), as predicted by H3b and H3c, respectively. Finally, communicative responsiveness was negatively related to inefficacy (β = −0.18, p < 0.01), in support of H3d. In addition to these hypotheses, the Final Model also supported the following: (1) a negative relationship between communicative responsiveness and cynicism (β = −0.11, p < 0.01); (2) a positive relationship between surface acting and inefficacy (β = 0.16, p < 0.01); and (3) negative relationships between automatic regulation and (a) exhaustion (β = −0.22, p < 0.01) and (b) cynicism (β = −0.30, p < 0.01). Contrary to H4a and H4b, deep acting was not significantly related to either exhaustion or inefficacy.

The foregoing results provide preliminary evidence in support of the hypothesized indirect relationships between professional seniority and one or more of the three dimensions of burnout via automatic regulation (H5a) and surface acting (H5d and H5e), but not via deep acting (H5b and H5f) or communicative responsiveness (H5c). Therefore, hypotheses 5b, 5c, and 5f are not supported by the data and the significance of those hypothesized indirect pathways was not further explored. The Final Model also suggested unpredicted indirect pathways between professional seniority and (1) inefficacy via surface acting; (2) cynicism via automatic regulation; and (3) exhaustion via automatic regulation. We tested the significance of the foregoing preliminarily supported indirect effects using the model indirect command in Mplus in conjunction with the bootstrapping procedures recommended by Preacher and Hayes [43]. Standardized indirect effects were computed for each of 1000 bootstrapped samples. Ninety-five percent confidence intervals (95% CI) were computed by determining the indirect effects at the 2.5th and 97.5th percentiles.

Based on these statistical tests, professional seniority had negative indirect impacts on exhaustion (standardized b = −0.03, p < 0.001, 95% CI = −0.045, −0.014)) and inefficacy (H5a; standardized b = −0.05, p < 0.001, 95% CI = −0.069, −0.023)) via automatic regulation. While professional seniority did not have a direct relationship with cynicism (see above), it did negatively influence cynicism indirectly via surface acting (H5d; standardized b = −0.04, p < 0.001, 95% CI = (−0.057, −0.017)) and automatic regulation (standardized b = −0.04, p < 0.01, 95% CI = (−0.058, −0.018)). Finally, professional seniority had similar indirect negative effects on exhaustion (H5e; standardized b = −0.03, p < 0.001, 95% CI = (−0.051, −0.016)) and inefficacy (standardized b = −0.02, p < 0.001, 95% CI = (−0.033, −0.008)) via surface acting. Based on the foregoing, more seasoned attorneys reported less surface acting and more automatic regulation than less experienced attorneys. They also reported less burnout, both directly and indirectly via surface acting and automatic regulation.

Though not hypothesized, it is important to report the following zero-order correlations as they may reveal the reasons for which seniority in the practice of law influences emotional communication and, ultimately, burnout (also see Table 2). Senior attorneys reported significantly less workload (r = −17, p < 0.01), role conflict (r = −0.22, p < 0.01), and supervisor support (r = −0.44, p < 0.01) than junior attorneys. Unexpectedly, the group of seasoned attorneys also reported less familial (r = −0.12, p < 0.01) and coworker (r = −0.06, p < 0.05) support. Lastly, more senior attorneys reported stronger job control (r = 0.37, p < 0.01) as compared to those more junior.

4. Discussion

The aim of the current study was to extend knowledge about the effects of professional seniority on the three dimensions of burnout (cynicism, inefficacy, and exhaustion) as mitigated by work-related emotional communication, in furtherance of attorney sustainability. Data collected from attorneys confirm that more senior attorneys are better able to stave off the effects of burnout by managing their communication of emotion in the practice of law. In this section, we break down these findings and identify theoretical as well as practical implications that may affect the human sustainability of attorneys in their practice of law, as well as the long-term sustainability of their employing organizations.

Previous research confirms that burnout is a serious threat to worker sustainability. Managing one’s emotions in providing professional services can be an essential element of attorneys’ work to gain their clients’ trust and compliance. While emotion management may be required of all attorneys, over time these legal professionals appear to learn how to communicate to not only serve their clients, but also to better protect themselves from stressors that can lead to burnout. Through the years, several models have identified various types of communication and their effects on service workers. Our study finds a relationship between tenure in the practice of law and attorneys’ ability to reduce their burnout via communication.

Specifically, professional seniority had direct and indirect negative effects on exhaustion and inefficacy, with the indirect effects mediated by automatic regulation. Seniority alone did not affect whether attorneys experienced cynicism, but more senior attorneys did tend to experience less cynicism via automatic regulation. In other words, level of experience among the attorneys appeared to protect them from experiencing all three dimensions of burnout due in large part to their increased practice of automatic regulation. Thus, attorneys conveying the expected emotions for their work that matched their own emotions (regulated only in magnitude) was an effective way of reducing burnout.

We also found that the more senior group reported less exhaustion, cynicism, and feelings of inefficacy due, in part, to their lower levels of surface acting. Previous research has shown that surface acting is an especially taxing communicative function [26]. Senior attorneys’ comparatively lesser use of surface acting than that of junior attorneys may be due to more senior attorneys feeling less compelled to express artificial feelings to manage, persuade, or soothe their clients and coworkers.

Notably, although communicative responsiveness negatively influenced cynicism and inefficacy, no significant difference was found between senior and junior attorneys in their reported use of communicative responsiveness. There was also no significant difference in deep acting based upon professional seniority. These results indicate that duration of practice experience has no influence on lawyers’ engagement in deep acting or the empathic communication required by communicative responsiveness. Perhaps, as similarly theorized by Powers and Myers [18], the advocacy and representative nature of the practice of law attracts individuals with a particular aptitude for the expressions of empathy and internalization of emotion required for communicative responsiveness and deep acting, respectively; and that aptitude remains relatively stable over time. A particular aptitude for deep acting may also explain our findings of no significant relationship between deep acting and any of the dimensions of burnout. Simply, deep acting might not be as strain-inducing for those attracted to the legal profession as found for other professions (see also [18]).

4.1. Theoretical Implications

These findings suggest that future burnout research and theoretical models should account for participants’ time in their professions. Although we found this seniority-related difference among attorneys, it is still unclear whether certain qualities of their occupation are the primary factors in the relationships we found. For example, do attorneys’ obligations to zealously advocate for their clients in the representation of their clients’ positions and interests drive their greater reliance on automatic regulation than surface acting as they gain more professional seniority? Additionally, is the lack of any effect related to deep acting in the Final Model (see Figure 2) specific to only attorneys, who are paid to act and argue on behalf of others? We theorize that communication with clients has the most significant effect on burnout, but workers may be just as likely to engage in surface acting, deep acting, and automatic regulation with coworkers, supervisors, or other persons with whom their work requires them to interface. Thus, it remains unclear whether similarly tenured non-attorney professional service workers (e.g., financial planners, engineers, or consultants) would show similar patterns because their levels of engagement with clients, coworkers, and supervisors may vary from that of attorneys.

The current study also contributes to theory about how workers’ communication affects their own well-being. Hochschild [13] and other studies (e.g., [34,44]) have shown that emotional labor and emotional work contribute to the strain of work. We found, however, that some work-related emotional communication can moderate strain-induced burnout. Specifically, attorneys who favor automatic regulation over surface acting in their practice of law experience reduced levels of all three dimensions of burnout, the practical implications of which will be addressed below.

4.2. Practical Implications

While the above describes some of the theoretical implications to emerge from the study, we can also identify practical implications that can further promote human sustainability at work. First, professional seniority is no panacea for burnout. Experience alone may help professional service workers to gain confidence in the efficacy of their work and to suffer less exhaustion, but professional seniority in conjunction with appropriate regulation of emotional communication markedly reduced our study participants’ experiences of burnout across all three dimensions. The ability to better use emotion management in providing professional services is not in the exclusive purview of more senior professionals. Management should provide early, preemptive guidance to professional service workers about the significant role of emotional communication in their work. Attorneys and other types of professional service professionals should be trained on the long-term contribution of expressing inauthentic emotions to experiencing burnout. Management may perceive this as counter to their objectives of providing outstanding customer service and maintaining client relationships. However, management may be otherwise persuaded when they understand the long-term implications for employee well-being, retention, and sustainability.

Second, programs that educate and train individuals for professional service occupations should include sections in their curriculum that guide students about the effects of emotional communication specific to their industries. Awareness of these processes may cause them to be more mindful about their workplace communication and influence their long-term ability to sustain themselves in their chosen occupations. Third, but related to this, professional associations should advise their members to promote sustainability within their profession. This is particularly true for women. The demographics of participants in our groups resemble that of national data indicating that women lawyers disproportionately leave the private practice of law within their first 10 years in the profession for a myriad of work-related reasons, meriting future inquiry (see [45]).

4.3. Limitations and Future Research

While this study has many strengths, it also has limitations. First, and as in many other studies, our findings were based on reports from only one profession—here, practicing attorneys. Therefore, it is possible that these findings may not generalize to workers in all professions. While we believe many professional service employees may experience similar emotion-related engagement with clients and others in the workplace, we cannot make claims without additional research to enable comparisons. Second, this study’s sample was comprised of practicing attorneys licensed in one state in the southeastern United States who voluntarily participated in this study. Consequently, the results of this study may not generalize to attorneys practicing nationwide, although it is unlikely that emotional communication in the practice of law and its consequences systematically differ by state lines. It is also unknown whether those who opted into this study are inherently different from those who did not (e.g., experiencing more or less burnout). Thus, future research should seek to validate our findings in broader, more diverse samples of professional service providers and employ probability sampling strategies to improve generalizability.

Third, all study variables were self-reported, increasing social desirability concerns. However, this study’s focal variables were inherently internal and subjective, making objective observation and measurement difficult, if not inappropriate. For example, the target employee is in the best position to assess if he or she is faking (i.e., surface acting), internalizing (i.e., deep acting), or genuinely feeling (i.e., automatic regulation) emotions displayed to another. This holds for psychological strain perceptions as well. Thus, self-report measures were the most appropriate for the purposes of this study. Recognizing that questions of a sensitive nature (i.e., psychological ailments that might impact performance) might increase false reporting, we also sought to decrease this potentiality by administering our survey online to decrease privacy concerns and ensure confidentiality.

Finally, this study’s data were cross-sectional in nature and collected via a common instrument (i.e., online survey questionnaire). As a result, this study cannot be used to make causal or temporal claims. Furthermore, our results may overestimate the true relationships among this study’s variables due to common data assessment (i.e., common method variance). To assuage this concern, we included reverse-scored items in our questionnaire, examined the data for spurious response patterns, and conducted a Harman’s one-factor post-hoc test. The results of this test suggest that common method variance did not substantially impact the results of this study. That is, the total variance accounted for by one factor was 26.66%, far less than the value (50%) that indicates common method bias [46]. Nevertheless, future research should employ more complex designs to establish causal connections, temporal precedence, and further validity for the results of this study.

5. Conclusions

Ensuring that workers can sustain their health and well-being is a primary concern for professional service firms (PSFs). PSF professionals are especially at risk for burnout due to working long hours, carrying heavy workloads, and dealing with clients and coworkers in emotionally charged situations. This study focused on one subpopulation of PSF professionals—lawyers. With the high rate of lawyers exiting in the early years of their profession, this is an especially important sustainability concern. Law schools typically do not teach new lawyers about how to manage these stressors, but it appears that, over time, practicing lawyers learn how to cope with these demands. Experienced lawyers are better able to regulate their emotions and less likely to feel the need to express artificial emotions in dealing with others, better protecting them from burnout. We hope that this and other studies will inform law schools, law firms, and other PSFs on the value of training their employees on how to manage their emotions and interactions with others most effectively to sustain their health and well-being throughout the course of their careers.