Job Satisfaction of Department of Veterans Affairs Peer Mental Health Providers

Abstract

Objectives:

Department of Veterans Affairs (VA) peer specialists and vocational rehabilitation specialists are Veterans employed in mental health services to help other Veterans with similar histories and experiences. Study objectives were to (a) examine job satisfaction among these employees, (b) compare them to other VA mental health workers, and (c) identify factors associated with job satisfaction across the 3 cohorts.

Methods:

The study sample included 152 VA-employed peer specialists and 222 vocational rehabilitation specialists. A comparison group included 460 VA employees from the same job categories. All participants completed the Job Satisfaction Index (11 aspects and overall satisfaction ratings). Linear regression was used to compare job satisfaction and identify its predictors among the 3 cohorts.

Results:

Job satisfaction was fairly high, averaging “somewhat satisfied” to “very satisfied” in 6 (peer specialists) and 9 (vocational rehabilitation specialists) of the 11 aspects and overall job ratings. Adjusting for length of employment, age and gender resulted in no significant group differences with 2 exceptions: White peer specialists were less satisfied with pay and promotion opportunities than vocational rehabilitation specialists and comparison-group employees. Across all cohorts, shorter length of time employed in the job was associated with higher job satisfaction.

Conclusions and Implications for Practice:

The high job satisfaction levels among the 2 peer cohorts suggest support for the policy of hiring peer specialists in the VA. Furthermore, the results are consistent with those of the nonveteran samples, indicating that integrating peer providers into mental health care is possible in VA and non-VA settings.

Background

Peer providers are individuals with a history of mental illness who are trained to assist in the mental health recovery of others. The role of peer providers in mental health services is well recognized, and many states provide Medicaid reimbursements for their services (Eiken, Reuters, & Campbell, 2008). The Department of Veterans Affairs (VA) also recognizes the value of including peer providers in the care of Veterans with mental health problems. The VA began to fund full-time peer specialist (PS) positions in 2005. The requirement for all VA medical centers to offer individual or group services from PSs for Veterans treated for serious mental illness was announced in 2008 (U.S. Department of Veterans Affairs, 2008). The important role of PSs in VA mental health services was further recognized in President Obama’s Mental Health Executive Order Improving Access to Mental Health Services for Veterans, Service Members, and Military Families, which was announced in August 2012 (President’s Mental Health Executive Order, 2012). The order required the VA, by December 31, 2013, to hire 800 peer-to-peer support counselors to empower Veterans to support other Veterans and help ensure that their mental health care and overall service needs are met. The VA announced on November 5, 2013 that they exceeded this goal by hiring 815 PSs and peer apprentices (U.S. Department of Veterans Affairs, Office of Public and Intergovernmental Affairs, 2013). The PSs and peer apprentices are Veterans who have successfully dealt with their own mental health recovery for a minimum of 1 year. PSs are trained and certified, whereas peer support apprentices are undergoing training and certification to become PSs. The job description for PSs includes assisting Veterans with social skills; combating stigma; role modeling; informing Veterans about community supports; assisting Veterans in choosing jobs that match their strengths; overcoming job-related anxiety; and, importantly, sharing their recovery stories to model effective coping skills.

As part of a multipronged effort to end homelessness among Veterans, the VA nationally hired 400 Veterans, with histories of or at risk for homelessness, to work as vocational rehabilitation specialists (VRSs) providing community-based employment services to Veterans as part of the Homeless Veterans Supported Employment Program. They were recruited under a U.S. Office of Personnel Management regulation for hiring persons with a severe physical, cognitive, or emotional disability or history of such disability. Although not specifically identified as “peers,” they were required to have a current or history of disability as a condition for being hired under this initiative. Because these VRSs would work in mental health services and be involved in delivering vocational services to Veterans with mental health concerns, program leaders were strongly encouraged to recruit Veterans with histories of emotional disabilities. The vast majority of the VRSs hired had a history of psychiatric or substance-use disorders (C. Heitzmann, personal communication, March 26, 2015). The job description for VRS includes assessing occupational areas to target job search activities, job skills training, search strategies, and placement, helping Veterans develop community supports and getting the individual to accept help. Although the job activities differ between PSs and VRSs (Eisen et al., 2015), these two types of VA peer mental health providers share the same goal of helping other Veterans with similar histories and experiences.

The benefits of peer providers to individuals who receive services, to peer providers themselves, and to mental health service delivery systems have been documented (Solomon, 2004). The benefits for peer providers include provision of a structure to their day (Dunn, Wewiorski, & Rogers, 2008; Yuen & Fossey, 2003) and building job skills (e.g., creating a resume, interpersonal skills, and the confidence to maintain a job), which are transferrable to other future jobs (Mowbray, Moxley, & Collins, 1998; Rigby, 2007; Solomon, 2004). For example, G. Deegan (2003) and P. Deegan (1988) report using many of the tools that they teach to others to maintain their own wellness. For some peer providers, the job is a path to reenter the work world (i.e., a first step toward another job; Mowbray et al., 1998). Working with others helps rebuild social connections (Rigby, 2007; Yuen & Fossey, 2003). Being paid is certainly another important aspect of work (Mowbray et al., 1998; Yuen & Fossey, 2003), which can bring better living conditions, education, and quality of life (Dunn et al., 2008). Satisfaction of “being paid to be myself” has been reported, although complaints about the discrepancy in pay compared to other staff have also surfaced (Mowbray et al., 1998; Rigby, 2007).

Job satisfaction is one of the most studied attitudes in the organizational literature, with more than 10,000 published articles (Wright, 2006). Historically, the desire to understand workers’ job satisfaction stemmed from employers’ interest in maximizing performance and reducing costs associated with turnover or absenteeism, which are both inversely associated with job satisfaction (Muchinsky, 1990). In addition to reducing turnover and costs to employers, job satisfaction among workers with mental illness has been studied for its potential to provide information to assist workers to be successful (Resnick & Bond, 2001). In addition, a better understanding of job satisfaction and the factors that promote it during new, large-scale organizational initiatives, such as VA’s peer support implementation and the Homeless Veterans Supported Employment Program, can provide valuable guidance to employers, helping them shape these initiatives to maximize success. However, job satisfaction levels in peer support workers, particularly compared with other mental health professionals, have not been well studied. One study found a high job satisfaction level (averaging 4.69 on a question asking to rate satisfaction with overall job experience using a 5-point scale) among 27 PSs 1 year after their Certified Peer Specialist training (Salzer, Katz, Kidwell, Federici, & Ward-Colasante, 2009). Likewise, the 59 Certified Peer Specialists surveyed by Grant and colleagues averaged 3.45 (SD = 0.56) on a 4-point job satisfaction scale (Grant, Reinhart, Wituk, & Meissen, 2012). Of the six job satisfaction subscales, only pay had an average below 3 (M = 2.57, SD = 0.74). Another recent study surveying 100 members of the National Association of Peer Specialists found that greater role clarity and psychological empowerment were associated with higher job satisfaction levels (Davis, 2013). All of these studies were conducted on nonveteran PSs. The job satisfaction of VA peer mental health specialists and factors that are associated with their satisfaction remain unknown. Furthermore, implementation of peer providers in mental health services has raised concerns about stress inherent in the job and a potential for burnout within and outside of the VA (Morse, Salyers, Rollins, Monroe-DeVita, & Pfahler, 2012). Consequently, examination of job satisfaction and its correlates among these workers can help identify and prevent job-related stress and burnout.

In light of the previous findings among nonveteran peer mental health providers and the lack of similar research in the VA, the objectives of this study were to (a) examine job satisfaction in various aspects of the job for PSs and VRSs, (b) compare their job satisfaction with other similar mental health workers in the VA, and (c) identify factors associated with job satisfaction across the three cohorts. This research focus is timely given the recent increase in PS staff hired in VA and non-VA mental health services. Our findings can potentially help VA and non-VA services to retain peer mental health providers by better understanding their levels of job satisfaction and its associated factors.

Method

Overview

This study was part of a national survey of VA-employed PSs and VRSs across 138 VA Healthcare Systems in 49 states (Eisen et al., 2015). The study was approved by the Institutional Review Board of the Edith Nourse Rogers Memorial Veterans Hospital in Bedford, Massachusetts. Details of the study design and procedures are reported elsewhere (Eisen et al., 2015).

Study Sample: VA-Employed PSs and VRSs

To recruit PSs and VRSs to our study, we first informed their supervisors and facility directors about the study and answered their questions. We then emailed a letter with a unique study ID and survey website link to all eligible PSs (n = 288) and VRSs (n = 378) employed at the start of the study in December 2011. The letter explained the voluntary and confidential nature of the study and shared all of the points pertaining to an informed consent. Enrolled participants were instructed to complete the survey outside of work hours. Nonrespondents were sent up to six email and/or telephone reminders. Participants received a $20 gift card.

One hundred fifty-nine PSs (55% of those eligible) and 231 VRSs (61%) responded to the survey. Seven PSs and nine VRSs completed less than 50% of the survey and were excluded from the analysis, leaving a final sample of 152 PSs and 222 VRSs. The two cohorts had similar histories of mental illness, disability, and substance abuse disorders (Eisen et al., 2015).

VA Employee Comparison Group

The comparison group was drawn from the VA employees who completed the VA All Employee Survey in 2012, the year that most baseline data for our national survey of VA-employed PSs and VRSs were collected. The VA All Employee Survey is administered annually to all VA employees across the country (Osatuke et al., 2012). It serves as an important feedback mechanism from employees to the organization on workforce satisfaction and perceptions of the workplace. The VA All Employee Survey includes reliable and valid measures, including the Job Satisfaction Index (Nagy, 2002), as well as demographics (age, gender, race, ethnicity), veteran status, supervisory level, VA years of service, type of workplace setting, and occupation.

Information available from the VA All Employee Survey was used to select a VA employee comparison group matching our sample. PSs and VRSs are categorized as “other” certified/licensed or nonlicensed professionals for the purposes of the VA All Employee Survey. Thus, we selected a comparison group from these two groups of VA employees who also reported that they worked in inpatient, outpatient, or extended-care mental health services. We excluded psychiatrists, psychologists, nurses, and social workers from the comparison group because these professions require very different, specialized, and considerably more extensive training than “other” licensed and nonlicensed professionals. “Other” certified/licensed or nonlicensed professionals also include health technicians and licensed professional mental health counselors. The job descriptions for these positions include providing individual assessment and case management, psychotherapy, group education, counseling, and follow-up services as well as coordinating care including serving as an advocate on behalf of Veterans to ensure comprehensive service delivery and linking and referring Veterans to various VA health-care facilities. We further limited the comparison group sample to those employed at the VA for less than 5 years to match the maximum length of employment of the VA PSs and VRSs in our study sample. Tenure (years of VA employment) was previously found to be associated with job satisfaction using the VA All Employee Survey data (Teclaw, Osatuke, Fishman, Moore, & Dyrenforth, 2014). The comparison group included a total of 460 VA employees. Figure 1 presents a flowchart of the selection criteria and number of employees obtained at each selection stage.

Figure 1.

Flowchart of VA employee comparison-group selection criteria.

Study Data

The national survey of VA-employed PSs and VRSs included demographic characteristics, length of time in the job, and the Job Satisfaction Index, also used in the VA All Employee Survey. Our analysis included the following variables: age, gender, race, years employed at the VA, veteran status, and the Job Satisfaction Index scores, all of which were available in both datasets (our study survey and the VA All Employee Survey).

Study Measures

The 12-item Job Satisfaction Index assesses overall satisfaction and 11 specific satisfaction aspects: type and amount of work, pay, coworkers, direct supervision, senior management, opportunities for promotion, working conditions, customer satisfaction, praise, and quality of work (Nagy, 2002). A single-item approach is used to measure each aspect of job satisfaction. This approach was shown to be highly correlated with multiple-item measures, yet it has the advantage of being easier and quicker to complete (Nagy, 2002). The items are rated on a 5-point scale from 1 (not at all satisfied) to 5 (very satisfied).

Statistical Analysis

To examine job satisfaction differences among the three cohorts (PSs, VRSs, and the comparison group of VA employees), we first compared mean item ratings of each cohort using a regression model. The model included an indicator for study cohort as an independent variable without adjusting for demographics or length of employment. We then identified demographics and length of employment that might explain differences in job satisfaction among the cohorts using regression models. Only aspects of job satisfaction that were found to have significantly different unadjusted ratings among the three cohorts were included in this analysis. We first examined the associations between demographics and length of employment with each aspect of job satisfaction within each cohort using separate linear regression models for each cohort. On the basis of the results of these regression models, we determined which variables to include in a regression model that used data from all three cohorts. This regression model was used to compare job satisfaction among the three cohorts adjusting for demographics and length of employment that were found to be associated with aspects of job satisfaction in each cohort. For demographic variables and length of employment that had similar associations with job satisfaction across cohorts, we only included the main effect variables in the regression model. For demographic variables and length of employment that had different associations with job satisfaction across cohorts, we also included the interaction terms among demographic variables, length of employment, and cohort indicators in the regression model.

Adjusted mean ratings for each aspect of job satisfaction were then calculated using the regression coefficients obtained from the regression model. The p values of paired cohort comparisons on the adjusted mean ratings were then also calculated.

Results

Demographic Characteristics and Length of Employment (Table 1)

Table 1.

Sample Characteristics

	PSs (na = 152)		VRSs (na = 222)		Comparison- group VA employees (na = 460)
Characteristic	n	%	n	%	n	%	x2
Age (years)							106.5**** (df = 6)
<40	13	8.7	29	13.3	181	39.6
40–49	36	24.2	60	27.5	123	26.9
50–59	73	49.0	102	46.8	98	21.4
>=60	27	18.1	29	12.4	55	12.0
Gender							66.9**** (df = 2)
Male	121	80	162	73	220	48
Female	31	20	60	27	237	52
Race							42.0**** (df = 2)
White	86	56.6	121	54.5	352	76.5
Non-White (majority African Americanb)	66	43.4	101	45.5	108	23.5
Veterans							192.3**** (df = 2)
Yes	133	88.7	215	98.6	232	50.7
No	17	11.3	3	1.4	226	49.3
Time in job							319.4**** (df = 6)
<6 months	29	19.5	31	14.0	24	5.3
6–12 months	7	4.7	156	70.3	87	19.1
1–2 years	26	17.5	33	14.9	84	18.4
2–5 years	87	58.4	2	0.9	261	57.2

^aThe number of participants included in the analysis for each variable varied slightly due to missing data on a few variables.

^bIncluding 4 American Indians in PS cohort, 2 American Indians and 3 Asians in the VRS cohort, and 3 American Indians and 11 Asians in the comparison VA employee cohort.

^****p < .0001.

The demographic characteristics of the comparison group of VA employees were significantly different from VA PSs and VRSs in several ways (all ps < .0001). Comparison-group VA employees were younger than PSs and VRSs, with 40% younger than 40 years old, whereas nearly half of PSs and VRSs were between 50 and 59 years old. Comparison-group VA employees had approximately equal numbers of males and females, whereas only 20% and 27% of PSs and VRSs were female. More than three quarters of the comparison group VA employees were White, yet only slightly more than half of PSs and VRSs were White. The veteran status of the comparison group also differed significantly from PSs and VRSs. Half of the comparison-group VA employees were Veterans, whereas most PSs (89%) and nearly all VRSs (99%) were Veterans. The tenure of VA employment was also significantly different among the three groups. Nearly 60% of the comparison-group VA employees had been in the job for 2–5 years and only 5% were employed for less than 6 months. Fifty-eight percent of PSs had also been in the job for more than 2 years, and 20% had only been in the job for less than 6 months. Most (70%) VRSs had been in the job for 2–6 months.

Unadjusted Job Satisfaction Ratings of Three Study Cohorts (Table 2)

Table 2.

Unadjusted Job Satisfaction Item Ratings

	PSs		VRSs		Comparison group VA employees
Job Satisfaction items	M	SD	M	SD	M	SD
Type of work	4.42	1.00	4.39	0.94	4.34	0.93
Amount of work	4.20	1.07	4.21	1.03	4.03	1.06
Pay****	2.98	1.37	3.64	1.26	3.47	1.25
Relationships with coworkers	4.23	1.12	4.29	0.97	4.15	1.05
Quality of direct supervision***	4.07	1.33	4.35	0.96	3.97	1.26
Quality of senior managers****	3.58	1.36	3.97	1.14	3.45	1.27
Opportunities for promotion****	2.44	1.44	3.42	1.27	2.83	1.32
Working conditions	3.89	1.31	4.04	1.05	3.87	1.11
Perceived customer satisfaction	4.22	0.98	4.16	0.83	4.12	0.87
Amount of praise received***	3.93	1.24	4.10	1.05	3.71	1.22
Quality of your work	4.55	0.75	4.41	0.76	4.53	0.69
Overall satisfaction****	4.26	0.94	4.37	0.88	4.02	0.97

Note. The Job Satisfaction Index ratings range from 1 (not at all satisfied) to 5 (very satisfied). Significance level of overall group difference.

^***p < .001.

^****p <.0001.

The three study cohorts had significantly different unadjusted ratings in overall job satisfaction (p < .0001) and in five specific aspects of the job: pay (p < .0001), quality of direct supervision (p < .001), quality of senior managers (p < .0001), opportunities for promotion (p < .0001), and amount of praise received (p < .001). The PSs and VRSs had fairly high job satisfaction, averaging somewhat satisfied (rating of 4) to very satisfied (rating of 5) in 6 (PSs) and 9 (VRSs) of the 11 aspects and overall job ratings. PSs and VRSs had significantly higher overall satisfaction scores than the comparison group of VA employees. VRSs had the highest scores in all five aspects of the job listed above. The comparison group of VA employees had the lowest scores in quality of direct supervision, quality of senior managers, and amount of praise received. PSs had the lowest scores in pay and opportunity for promotion.

Factors Associated With Job Satisfaction (Table 3)

Table 3.

Regression Analysis Results of Selected Job Satisfaction Items

	Pay		Quality of direct Supervision		Quality of senior managers		Opportunities for promotion		Amount of praise received		Overall satisfaction
Independent Variables	ba	SEb	b	SE	b	SE	b	SE	b	SE	b	SE
Sex
Female	.05	.10	.06	.09	−.07	.10	−.24*	.10	.02	.09	−.09	.07
Malec	—		—		—		—		—		—
Age (years)
<40	−.43**	.16	−.18	.15	−.30	.16	−.06	.16	−.09	.15	−.23	.12
40–49	−.31*	.15	−.31*	.14	−.22	.15	−.13	.15	−.15	.14	−.18	.11
50–59	−.18	.15	−.07	.14	−.04	.14	−.15	.15	−.02	.14	−.09	.11
>=60c	—		—		—		—		—		—
Race
Non-White	.11	.17	.04	.16	.03	.17	.22	.17	.02	.16	.01	.13
Whitec	—		—		—		—		—		—
Employment length	−.18***	.05	−.25****	.05	−.27****	.05	−.38****	.05	−.26****	.05	−.19****	.04
Cohort
PS	−.76****	.19	−.20	.18	−.29	.18	−.89****	.19	−.08	.17	−.12	.14
Comparison group	.22	.15	−.06	.14	−.07	.15	.03	.16	−.06	.14	−.08	.11
VRSc	—		—		—		—		—		—
Cohort_race
PS non-White	.62*	.27	.38	.26	.35	.27	.71**	.27	.38	.25	.43*	.20
PS Whitec	—		—		—		—		—		—
AES non-White	−.42*	.22	−.04	.21	−.15	.22	−.28	.22	−.09	.20	.03	.16
AES Whitec	—		—		—		—		—		—
VRSc non-White	—		—		—		—		—		—
VRSc Whitec	—		—		—		—		—		—

Note. Selected job satisfaction items are the overall satisfaction and the five aspects of job satisfaction that had significantly different unadjusted scores among the three study groups. Employment length is the length of VA employment, an ordinal scale used as a continuous variable in the regression model. AES = comparison-group VA employees identified from the All EmployeeSurvey; PS = Peer Specialists; VRS = Vocational Rehabilitation Specialists.

^aRegression coefficient.

^bStandard error (SE).

^creference group.

^*p < .05.

^**p < .01.

^***p < .001.

^****p < .0001.

To understand what might have contributed to the differences among the three cohorts on the overall and specific aspects of job satisfaction previously described, we examined factors that were associated with Job Satisfaction Index item ratings. Within each of the three study cohorts, there were similar associations between job satisfaction item ratings and length of employment, sex, and age. However, significant differences between race categories (White vs. non-White) on job satisfaction ratings occurred in the PS cohort. Therefore, we included the interaction between race and cohort in addition to the main effects of age, sex, length of employment, cohort, and race in the regression model that used data from all three cohorts. The results of the regression analysis showed that for all three cohorts, length of employment had a strong negative association (p < 0.0001) with overall job satisfaction and with the five aspects of job satisfaction described above (those which had significantly different unadjusted ratings among the three study cohorts). The negative association indicated that the longer the employment, the lower the job satisfaction. Females were less satisfied than males in opportunities for promotion (p < .05) for all three study cohorts. Younger employees were less satisfied with pay than older employees for all three cohorts. Non-Whites had higher levels of satisfaction in pay and opportunities for promotion than Whites, but only for the PS cohort. There were no race differences among VRSs and comparison group employees in overall job satisfaction or any specific aspect of job satisfaction.

Adjusted Job Satisfaction Ratings of Three Study Cohorts (Table 4)

Table 4.

Adjusted Scores of Selected Job Satisfaction Items by Cohort and Race

Job Satisfaction Items
	PSs		VRSs		Comparison Group VA Employees
	M	SE	M	SE	M	SE
Pay***
White	2.67a	0.14	3.43b	0.13	3.65b	0.07
Non-White	3.40b	0.16	3.54b	0.14	3.34c	0.13
Quality of direct supervision
White	3.93	0.13	4.12	0.12	4.06	0.07
Non-White	4.34	0.15	4.16	0.13	4.06	0.12
Quality of senior managers
White	3.39	0.14	3.68	0.13	3.61	0.07
Non-White	3.77	0.16	3.71	0.14	3.49	0.12
Opportunities for promotion***
White	2.05d	0.14	2.94e	0.13	2.97e	0.07
Non-White	2.99e	0.17	3.16e	0.14	2.90e	0.13
Amount of praise received
White	3.80	0.13	3.88	0.12	3.81	0.07
Non-White	4.20	0.15	3.90	0.13	3.74	0.12
Overall satisfaction
White	4.06	0.10	4.18	0.09	4.10	0.05
Non-White	4.50	0.12	4.19	0.10	4.14	0.09

Note. The Job Satisfaction Index score is adjusted for age, sex, and length of employment. Ratings range from 1 (not at all satisfied) to 5 (very satisfied).

a is significantly lower than b, p < .0001; a is significantly lower than c, p < .0004; b and c are not significantly different; d is significantly lower than e, p < .0001.

Interaction between cohort and race: ***p <.001.

Demographic characteristics, length of employment, and the interaction terms that were associated with job satisfaction as described earlier were used as adjusting variables for comparing cohorts on overall satisfaction and on the five specific aspects of job satisfaction that showed statistically significant unadjusted differences among the cohorts. Two aspects of job satisfaction—pay and opportunities for promotion—showed significant differences among the study cohorts after adjusting for age, sex, and employment length, but these differences only occurred in Whites (see Table 4). White PSs had significantly lower levels of satisfaction than White VRSs and White comparison-group employees in pay and in opportunities for promotion (p < 0.0001). After adjusting for age, sex, and employment length, there were no longer significant group differences in overall job satisfaction, quality of direct supervision, quality of senior managers, or amount of praise received.

Discussion

Our results show that after adjusting for length of employment, age, and gender, no differences in job satisfaction were found among PSs, VRSs, and other similar VA employees working in mental health direct-care positions with two exceptions: White PSs were less satisfied with pay and opportunities for promotion than both VRSs and comparison-group employees. Across all three cohorts, length of employment was a strong predictor of job satisfaction. Employees with shorter tenure had higher levels of job satisfaction than employees with longer tenure. Thus, it is likely that the unadjusted higher levels of job satisfaction for peer and VRSs might mainly be explained by their shorter length of employment even within the 5-year length of the employment cap, which we imposed for comparability purposes.

The lower levels of satisfaction with pay and promotion opportunities observed among White PSs might reflect the lower job grade level of the VA PS employees. We do not have salary data for the VA comparison group, but at the time of this study, the pay grades for PSs were lower than those for the VRSs. However, the differing pay grades would not account for the fact that only White PSs were less satisfied with pay and promotion opportunities than VRSs and comparison-group employees. The VA has recently upgraded the pay for PS positions, which may decrease or eliminate differences in satisfaction with pay and promotion found among White PSs.

Regarding race differences in job satisfaction, inconsistent results have been reported in the literature. A recent study using two national surveys—the General Social Survey and the National Longitudinal Survey of Youth—found that Blacks were significantly less satisfied with their jobs than Whites and that perceived discrimination, rather than actual labor market discrimination, explains part of the difference (Mukerjee, 2014). To the contrary, and consistent with our findings, an earlier study using data from the National Longitudinal Surveys of Mature Men found that older Black, male adults were more satisfied with their jobs than older White adults (Bartel, 1981). These studies only assessed a single item of overall job satisfaction. Our study found that in the PS cohort, non-Whites (94% of which were Blacks) were more satisfied with pay and opportunities for promotion. Future studies are warranted to further examine the race difference and possible reasons for the difference.

The peer and VRSs had similar job satisfaction levels as comparison group employees in all other aspects of the job as assessed by the Job Satisfaction Index after controlling for demographic characteristics. However, differences might have been missed. In the VA comparison group cohort, those who were Veterans had lower job satisfaction than non-Veterans (data not shown). Because almost all of the PSs and VRSs were Veterans, if we could have controlled for veteran status in the regression analysis, differences between PSs and VRSs with comparison group employees would have been larger. Higher levels of satisfaction among VA PSs and VRSs compared with other mental health providers could derive from the flexibility to use and share their identity as Veterans with histories of mental illness, substance abuse, and/or homelessness in their work with Veterans receiving mental health services. Comparison-group employees are much less likely to be Veterans or to have experienced these conditions themselves. Following a more traditional model for mental health providers, they are probably less likely to share elements of their personal histories or experiences with the Veterans they treat. Qualitative findings from other peer samples find that peer providers refer to their job as “a calling” and “a mission” and note that being a peer provider made their lives more “complete” and “fulfilling” (Moran, Russinova, Gidugu, Yim, & Sprague, 2012). This strong sense of mission to help similar individuals could likely contribute to job satisfaction.

One limitation of this study is the unknown overlap between the VA All Employee Survey sample and the PS and VRS samples. It is possible that some of the employees in the PS and VRS samples also completed the VA All Employee Survey in 2012, but it is not possible to separate them from the rest of the comparison-group sample because the VA All Employee Survey is anonymous. Thus, we cannot determine if some of our survey respondents also responded to the VA All Employee Survey in the same year. On the basis of the numerous demographic differences between the VA All Employee Survey sample and the PS and VRS samples (e.g., with respect to age, gender, veteran status, and race), substantial overlap is unlikely. However, even if it were present, an overlap would result in underestimating the differences between the PS and VRS samples and the broader sample of mental health staff who completed the All Employee Survey.

It might also be possible that the comparison group included PSs and VRSs who were not in our two study cohorts. However, because 55% and 61% of all PSs and VRSs, respectively, responded to our survey, which is close to the 63.4% response rate of 2012 VA All Employee Survey (U.S. Department of Veterans Affairs, VHA National Center for Organization Development, 2013), and our survey provided monetary incentive to the Veterans, which the All Employee Survey did not, the probability that PSs and VRSs would complete the All Employee Survey, but not our survey, is likely very low. As such, the Veterans in the comparison group are most likely not PSs or VRSs.

One other study limitation is that comparison-group employees who completed the All Employee Survey might not be representative of all VA employees in similar mental health provider positions. The VA All Employee Survey, 2012 had a relatively high response rate, 63.4%, which represents voices from more than half of all VA employees. However, we recognize that this high overall response rate might not necessarily apply to our comparison group, which represents a small proportion of all of the VA employees who completed the All Employee Survey, 2012 (Figure 1). Yet another limitation is that the Job Satisfaction Index used in the study does not include some job aspects possibly relevant to peer mental health providers—such as role conflict, mixed messages about socialization with colleagues, and lack of clarity about health information privacy, which might affect the job satisfaction of peer providers (Gates & Akabas, 2007). Future studies need to examine these aspects of job satisfaction.

Finally, there are no survey data that directly support our assertion that the VRSs were mental health peers. However, in addition to VA leader’s assurance that they were mental health peers, our data also indicate that the VRS cohort had similar histories of mental illness, disability, and substance abuse disorders as the PS cohort (Eisen et al., 2015).

To conclude, our study is the first to examine the job satisfaction of peer mental health providers in the VA. Our findings that the PS and VRS VA employees had fairly high job satisfaction are similar to the findings on nonveteran PSs reported in other studies (Davis, 2013; Grant et al., 2012; Salzer et al., 2009). In addition to corroborating these previous findings, our study provides further information about the comparison of peer providers and nonpeer providers in the VA. It is encouraging to find little difference in job satisfaction between VA peer mental health providers and other VA employees in similar direct-care mental health positions. However, because of the limitations of the data available for deriving the VA employee comparison group, this finding warrants further examination and replication.

Our findings of high levels of job satisfaction among veteran PS and VRS employees have important implications. First, these results suggest empirical support for the policy of hiring PSs in VA mental health services outlined in the recent executive order (President’s Mental Health Executive Order, 2012). Second, the findings address a widely acknowledged concern regarding the ability of civilian organizations to adequately support veteran employees in transitioning from military to civilian jobs. Veterans share several unique needs and strengths, which are often not well understood by civilian employers (King, 2011). These unique characteristics, along with a dearth of information regarding how their strengths can be leveraged in the civilian workplace, put veteran employees at risk for unemployment and underutilization of their skills in civilian organizations. Our findings, based on the data of two specific occupations in the VA, suggest that success in military personnel transitions to civilian jobs is possible. One reason for this success could be that the occupations in question leveraged the unique strengths brought to the workplace by these Veterans—a shared experience with those for whom they were providing care. Although these results within the VA, the largest federal employer organization in the United States, are highly encouraging, we acknowledge that they need to be replicated in subsequent studies using larger and more diverse samples as well as research designs that overcome some limitations noted for the current study.