Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Oct 1.
Published in final edited form as: J Subst Abuse Treat. 2010 Aug 1;39(3):264–271. doi: 10.1016/j.jsat.2010.06.009

How Serious of a Problem is Staff Turnover in Substance Abuse Treatment? A Longitudinal Study of Actual Turnover1

Lillian T Eby 1, Hannah Burk 2, Charleen P Maher 3
PMCID: PMC2937083  NIHMSID: NIHMS219624  PMID: 20675097

Abstract

In the substance abuse treatment field, the annual turnover rate is cited as being anywhere between 19 and 50 percent (Johnson & Roman, 2002; Gallon, Gabriel, & Knudsen, 2003; Knudsen et al., 2003; McLellan et al., 2003). However, no research to date has evaluated these claims by tracking turnover longitudinally using organizational turnover data from substance abuse treatment centers. This research presents the results of a longitudinal study designed to systematically examine actual turnover among counselors and clinical supervisors. Twenty-seven geographically dispersed treatment organizations, serving a wide range of clients in the public and private sector, provided data for the study over a two year time span (2008–2009). The annual turnover rate was 33.2% for counselors and 23.4% for clinical supervisors. For both groups the majority of turnover was voluntary (employee-initiated). Specific reasons for turnover were largely consistent across the two groups, with the most common reason being a new job or new opportunity. The findings are discussed in terms of the unique employment context of substance abuse treatment. Practical recommendations are also discussed to help stem the tide of turnover in the field of substance abuse treatment.

Clinician turnover is discussed as a major problem in substance abuse treatment facilities (Ducharme, Knudsen, & Roman, 2008; Knight, Bromme, Simpson, & Flynn, 2007; Knudsen, Johnson, & Roman, 2003; McLellan, Carise, & Kleber, 2003), with estimates of annual turnover rates ranging between 19% and 50% (Johnson & Roman, 2002; Gallon, Gabriel, & Knudsen, 2003; Knudsen et al., 2003; McLellan et al., 2003). Despite the claim that turnover is a serious concern we are aware of no published studies that systematically track the employment status of individuals over time to ascertain actual turnover. Rather, some existing research used program administrator estimates of clinician turnover rates (Gallon et al., 2003; Knudsen et al., 2003; Johnson & Roman, 2002) which may or may not be accurate. Other estimates of turnover were from larger research projects which do not specify how information on clinician turnover were obtained or calculated (Carise, Gurel, McLellan, Dugosh, & Kendig, 2005; Carise, McLellan, Gifford, & Kleber, 1999; McLellan et al., 2003).

Based on the current state of research, the actual rate of clinician turnover and the reasons why individuals leave substance abuse treatment organizations remain elusive. The present study pursues three goals to address this gap in the literature. First, using longitudinal data we track actual turnover among a geographically dispersed sample of full-time counselors and clinical supervisors working in community-based substance abuse treatment organizations throughout the United States. This yields information on both voluntary (employee-initiated) and involuntary (employer-initiated) turnover, which is an important distinction since voluntary turnover is generally viewed as undesirable for the organization, whereas involuntary turnover is typically viewed as less undesirable. Second, we provide an in-depth account of the specific reasons why counselors and clinical supervisors leave their employing organizations (both voluntarily and involuntarily) by content analyzing qualitative data on the reason for departure derived from organizational records. Third, turnover rates and reasons for turnover are examined separately for counselors and clinical supervisors. Finally, we use interview data from a sub-sample of former employees to verify the accuracy of the organizational data and to provide insight into post-turnover employment decisions.

Importance of Clinician Turnover in Substance Abuse Treatment

As with any organization, both voluntary and involuntary clinician turnover incurs a direct financial cost in terms of recruitment, selection, and training (Alexander, Bloom & Nuchols, 1994; Staw, 1980). Turnover can also reduce organizational efficiency (Alexander et al., 1994; Kacmar, Andrews, Van Rooy, Steilberg, & Cerrone, 2006), particularly if high-performing employees leave. Organizations with high rates of turnover also often suffer from low employee morale (Johnson & Roman, 2002), which can have reverberating negative effects on the organization. Indirect costs also accrue with turnover such as the loss of institutional knowledge and less success in the adoption and implementation of evidence-based practices for treating substance abuse (Carroll & Rounsaville, 2007; Saxon & McCarty, 2005). Finally, turnover can increase stress on remaining counselors by increasing their caseloads in an effort to meet treatment demands (Knight et al., 2007; Powell & York, 1992).

Turnover can also have a deleterious effect on patient care (Ducharme et al., 2008; Lum, Kervin, Clark, Reid, & Sirola, 1998). Research finds that patients stay in drug treatment longer if they have the same counselor (McCaul & Svikis, 1991), have more therapeutic contact with longer tenured counselors, and do better in drug treatment if there is continuity in treatment provision (Lamb, Greenlick, & McCarty, 1998). In addition, a stronger therapeutic alliance between counselor and patient is associated with greater drug treatment participation, more days abstinent, and fewer drinks per day (Connors, Carroll, DeClemente, Longabaugh, & Donovan, 1997). At the system level, there is the concern that the current national shortage of substance abuse counselors is a key barrier to expanding service delivery capacity [Center for Substance Abuse Treatment (CSAT), 2000; Whitter, Bell, Gammond, Gwaltney, Magana, & Moreaux, 2006]. Adding to this crisis are repeated reports of serious problems recruiting qualified staff who meet minimum job requirements (Gallon et al., 2003).

Review of Existing Research on the Turnover Problem in Substance Abuse Treatment

Probably the most widely cited estimates of turnover are from McLellan and colleagues’ research on efforts to improve the substance abuse treatment system. Carise, McLellan et al’s (1999) report on the development and initial use of an electronic information system (the Drug Evaluation Network Survey or DENS) included pilot data from a sample of 34 adult-only treatment programs in five urban cities. Eight of the 34 treatment programs (24%) dropped out of the pilot study due to turnover among program directors and administrators. Carise et al. (1999) also reported that on average, 50% of program staff either left or changed positions within a year. Unfortunately, it is not clear who reported this information or what sources of data are used to compile turnover estimates. Moreover, Carise et al. noted that their data are non-representative and that their results should be interpreted cautiously.

A commentary on the infrastructure crisis in the substance abuse treatment system is also cited as evidence of the turnover problem (McLellan et al., 2003). This study interviewed program directors and other staff members from a nationally representative sample of 175 substance abuse treatment programs drawn from the list of facilities in the 2000 edition of the National Survey of Substance Abuse Treatment Services (N-SSATS). This study focused on a wide range of structural and process characteristics of treatment programs, with an eye toward those that may hinder efficiency and could compromise treatment care or capacity. One conclusion noted by McLellan et al. is the “extreme instability of the workforce at all levels within the national treatment system” (p. 120). Although no specific statistics are provided and the data source is not clear, the article reported that over half of program directors had not been in their jobs for one year and similarly high rates of turnover were found among counseling staff.

Carise et al. (2005) provided a more specific estimate of clinician turnover based on a technology transfer study conducted in nine community-based drug treatment programs in Philadelphia. The authors experienced difficulty with initial recruitment, which program administrators attributed in part to high levels of staff turnover. Although not the focus of the study, these authors noted in the discussion section that over six months of the study, 32% of the counselors had left the treatment organization and another 27% dropped out of the study.

In another study Gallon et al. (2003) surveyed substance abuse treatment program directors and clinical staff working in four states in the Pacific Northwest. Although the response rate was rather low (43%), agency directors reported an average loss of 2.78 staff per year, which was estimated to reflect an annual turnover rate of nearly 25% since the average agency employed 11 staff members. Gallon et al. also provided some insight into the reasons for turnover. On average, treatment programs reported losing 1.75 staff members annually due to resignations, .36 due to layoffs, and .67 due to termination for performance.

Finally, research by Roman and colleagues provided information on the turnover rate of counselors using self-report data from program administrators. Johnson and Roman (2002) obtained estimates from 450 program administrators and reported an average counselor turnover rate of 24% (ranging from a low of 0% to a high of 50%). No data were provided on clinical supervisor turnover. Another study also used program administrator self-report data and found an average turnover rate of 18.5% for counselors based on data from the National Treatment Center Study (Knudsen et al., 2003).

Taken together, existing research has brought the issue of clinician turnover to the attention of researchers and practitioners. Unfortunately, there is wide fluctuation in estimates of clinician turnover, limiting our understanding of how serious the turnover problem is in substance abuse treatment. Furthermore, with the exception of Gallon et al. (2003), there is little concrete information as to why counselors and clinical supervisors are leaving their jobs. The present study addresses these gaps in the literature.

Method

Participants

Organizational representative

In 47% of the organizations turnover information was provided by human resource director or manager. In 32% of the organizations the COO/VP or president compiled the information (with input from clinical supervisors and/or HR). In the remaining organizations, 16% had clinical supervisors complete the form and 5% had an executive assistant complete the form (using HR records). The organizational survey was completed primarily by individuals occupying the title of Chief Executive Officer, Director, or Vice President of the 27 treatment organizations that comprised the sample.

Former employees

An employee survey provided information on the socio-demographic characteristics of counselors who turned-over in Year 2: 68% female, average age of 40 years, 58% Caucasian, modal education level of Masters or professional degree, and salary of $32,000. The socio-demographic profile of clinical supervisors is as follows: 53% female, average age of 43 years, 70% Caucasian, modal education level of Masters or professional degree, and salary of $37,633. Structured interview data were collected from a sub-sample of former employees; 66 counselors (27% of turnover sample) and 14 clinical supervisors (32% of turnover sample). The socio-demographic characteristics of this sub-sample did not differ significantly from the full sample.

Procedure

As part of a larger NIDA-funded research project, longitudinal data were collected from 86 geographically dispersed, free-standing public and private treatment centers that were part of 27 treatment organizations throughout the United States. Because the funded project used NIDA’s Clinical Trials Network (CTN) (http://www.drugabuse.gov/CTN/) as a platform for the research, all but one treatment organization was part of the CTN at the time of recruitment. Treatment organizations were recruited during formal presentations at the CTN’s External Affairs Subcommittee Meeting and the Community Treatment Program Caucus (both in May 2004). In addition, several community treatment program directors, all of whom were known leaders within the CTN, were briefed on the study and asked to help with study recruitment. It was through these efforts that 26 CTN affiliated treatment centers and 1 non CTN treatment organization agreed to participate. Importantly, there is evidence that counselors employed at programs affiliated with the CTN are similar to other counselors across the country (Knudsen, Ducharme, & Roman, 2007).

A treatment program is defined as a relatively autonomous, free-standing operational unit. Treatment organizations refer to larger organizational structures with oversight and authority over participating treatment centers. To participate in the study, eligible treatment organizations had to provide drug abuse counseling services in a community-based setting. This excluded prison-based programs, Veteran’s Health Administration programs, and driving-under-the-influence schools. Eligible counselors had to have direct contact with clients in a therapeutic relationship (individual or group counseling sessions, or both) and eligible clinical supervisors had to provide written and/or face-to-face clinical supervision to counselors. The research protocol was approved by the Institutional Review Board at the first author’s home university.

Organizational data

The information used to compile counselor and clinical supervisor turnover information were collected at two points in time, separated by a one year interval. We refer to these two waves of data as “Year 1” and “Year 2” from here forward. A representative from each of the 27 participating treatment organizations compiled turnover data and also completed a survey in Year 2 which provided information on the treatment organization. In 95% of the organizations the reason for turnover was formally documented in the employee’s personnel file. This information was obtained via an exit interview in 83% of the organizations. In all of these cases the former employee was specifically asked why he or she was leaving his or her job during the interview. In addition, 93% of the organizations that used an exit interview indicated that they asked a standard set of questions for employees. For those organizations that did not use exit interviews, they reported asking the employee directly, obtaining the information from employee resignation letters (if voluntary turnover), or using multiple methods (e.g., asking HR, asking employee directly, asking supervisor). Participating treatment organizations received compensation ($1000) at each data collection time point to off-set the staff time required to compile the data.

Employee data

As part of a larger research project, a confidential and voluntary paper-and-pencil survey was collected in group setting (separately for counselors and clinical supervisors). A research assistant proctored each session and organizational representatives were not present during survey completion. Employees turned in completed surveys to the research assistant who personally carried completed surveys back to the university for processing. The response rate was over 80% for both counselors and supervisors. Organizational response rates ranged from a low of 50% to a high of 100%. The data were collected on-site during normal business hours and the treatment center was compensated $50 for each completed counselor survey and $75 for each completed clinical supervisor survey to off-set employee time off-the-clock.

Following survey completion in Year 1, employees had the option to fill out a contact form to allow us to follow-up with them by email or phone if they left the treatment center between Year 1 and Year 2. Seventy-six percent of survey participants provided this information, which was kept separate from completed survey. Structured telephone interviews were conducted by a member of the research team. Interviews were recorded (with participant permission) and transcribed verbatim. Questions were asked concerning their departure from the organization. This allowed us to compare the information obtained by the organizational with that reported by a convenience sample of former employees. It also provided information regarding former employees’ current employment. Participants were compensated $40 for the interview.

Measures

Turnover

In Year 1, we obtained lists of all full-time counselors and full-time clinical supervisors employed at each treatment program from the program administrator. Our decision to focus only on turnover among full-time employees is because it is more costly to organizations since there is typically greater investment in the training and development of full-time employees. Moreover, full-time employees typically represent the “core” of an organization, occupying jobs that require greater skill, education, and experience, making full-timers more difficult to replace than part-timers.

A total of 739 counselors and 188 clinical supervisors were included on the Year 1 participant lists. In Year 2, we sent the Year 1 lists back to each program administrator and inquired about the employment status of each counselor and clinical supervisor on the Year 1 list. For individuals who were noted as no longer employed at the treatment program in Year 2, information was obtained from organizational records. We asked the program administrator to provide the following information on each employee on the Year 1 list: (1) still employed at treatment organization (yes or no), (2) if no longer employed, was individual terminated (yes or no), and (3) reason for turnover from the treatment organization (open-ended response from employee record). In 95% of the organizations a specific reason was formally documented in the personnel file. Eighty-three percent of organizations conducted actual exit interviews, and in 93% of the cases a pre-determined set of questions were asked of all employees. If an exit interview was not used the information was obtained by asking employees directly, obtaining information from supervisors, or culling information from resignation letters.

Employment-related questions

A series of standard questions were asked of former employees during the interview conducted by the research team. Of relevance to this study are questions regarding whether or not their departure from the organization was voluntary or involuntary, the specific reason(s) for leaving, whether or not they are currently re-employed, if re-employed whether or not they are working in the field of substance abuse treatment, and if not, why.

Content Analysis Procedure

Open-ended data on the reasons for turnover were content analyzed using the guidelines suggested by Miles and Huberman (1994). Specifically, two trained research assistants independently read through the open-ended responses provided as to the reason for turnover from the Year 2 lists. Next, each research assistant generated a list of potential coding categories to capture the different reasons for turnover. Agreement among coders on the initial coding categories was 90%. After discussion and consensus there was 100% agreement on the final coding categories.

Results

The treatment organizations were mostly accredited (72%), non-profit entities (88%) that were not housed on a hospital campus (84%). All major areas of the United States were represented in the study as follows: 26% Eastern US (e.g., Connecticut, New York), 11% Midwest US (e.g., Indiana, Ohio), 26% Southern US (e.g., Florida, North Carolina), and 37% Western US (e.g., California, Oregon). An average of 27 counselors and 7 clinical supervisors were employed at participating organizations. Included in the sample are treatment organizations that offered services to adolescents only, adults only, and both adolescents and adults. In-patient, residential, and out-patient programs were also represented in the sample. Most treatment organizations (60.9%) reported their treatment orientation as being eclectic or mixed model.

Reasons for turnover

The results of the content analysis yielded 12 voluntary turnover reasons: new job/other opportunity, job dissatisfaction, higher pay, other job-related (e.g., on-call), relocation, personal health, family, returned to school, retired, other personal, resigned, and voluntary, no reason provided. We also collapsed or clustered specific reasons for voluntary turnover into the following higher-order categories as recommended by Hycner (1985): job-related reasons (new job/other opportunity, job dissatisfaction, higher pay, other job-related) and personal reasons (relocation, personal health, family, returned to school, retired, and other personal). Two other higher-order categories of voluntary turnover are resigned and voluntary, no reason provided. Three specific involuntary turnover reasons emerged from the data: terminated, program closure/lay-off, and deceased. There were missing data on both the type and reason for turnover on 6 counselors and 3 clinical supervisors.

Counselor turnover

Turnover rates for treatment organizations varied widely. Three of the 27 treatment organizations experienced no counselor turnover and turnover ranged from 0% – 58% across organizations. The average annual counselor turnover rate was 33.2%. About three-quarters (75.1%) of those who left their treatment organization did so voluntarily, 22.4% left involuntarily, and the remaining 2.4% represented missing data. This translates into a voluntary turnover rate for counselors of 24.9% (33.2% × .751) and an involuntary turnover rate of 7.4% (33.2% × .224). As shown in Table 1, the most frequent higher-order turnover category was job-related (32.7%). The most commonly reported reason for job-related turnover was a new job or other opportunity (27.3%). An additional 19.2% of counselors who left the organization did so for personal reasons (see Table 1). The most common specific personal reason was relocation (6.9%), followed by personal health (5.3%). Of the remaining counselors who left involuntarily, the main reasons were termination (15.1%) and program closure/lay-off (6.9%).

Table 1.

Reasons for Counselor Turnover

N % n % n %
Voluntary Turnover 184 75.1%
 Job-Related Reasons 80 32.7%
  New job/other opportunity 67 27.3%
  Job dissatisfaction 6 2.4%
  Higher pay 5 2.0%
  Other job-related (e.g., on-call) 2 0.8%
 Personal Reasons 47 19.2%
  Relocation 17 6.9%
  Personal health 13 5.3%
  Returned to school 5 2.0%
  Other personal 5 2.0%
  Retired 4 1.6%
  Family 3 1.2%
 Resigned 35 14.3%
 Voluntary, no reason provided 22 9.0%
Involuntary Turnover 55 22.4%
 Terminated 37 15.1%
 Program closure/Lay-off 17 6.9%
 Deceased 1 0.4%
Missing data 6 2.4%
Open in a new tab

Total N=245. Percentages may not equal 100% due to rounding.

Clinical supervisor turnover

There was substantial variability in turnover rates across treatment organizations. Nine of the 27 treatment organizations experienced no clinical supervisor turnover and turnover ranged from 0% – 75% across organizations. The average annual turnover rate for clinical supervisors was 33.4%. Almost two-thirds (61.4%) left voluntarily, 31.8% left involuntarily, and there were missing data on 6.8%. Therefore, the voluntary turnover rate for clinical supervisors is 14.4% (23.4% × .614) and the involuntary turnover rate is 7.4% (23.4% × .318). As with counselors, the most commonly reported reason was job-related (36.4%, see Table 2). Within this higher-order category of job-related reasons, a new job or other opportunity (29.5%) was most often cited, followed by other job-related reasons (4.5%). Personal reasons accounted for an additional 13.6% of the turnover among clinical supervisors, with the most common specific reasons being retired (4.5%). Of the remaining clinical supervisors that left the organization involuntarily, 15.9% were terminated, 13.6% left due to program closure/lay-off and 6.8% reflected missing data.

Table 2.

Reasons for Clinical Supervisor Turnover

N % n % n %
Voluntary Turnover 27 61.4%
 Job-Related Reasons 16 36.4%
  New job/other opportunity 13 29.5%
  Other job-related (e.g., contract) 2 04.5%
  Higher pay 1 02.3%
  Job dissatisfaction 0 00.0%
 Personal Reasons 6 13.6%
  Retired 2 04.5%
  Relocation 1 02.3%
  Personal health 1 02.3%
  Family 1 02.3%
  Returned to school 1 02.3%
  Other personal 0 00.0%
 Resigned 2 04.5%
 Voluntary, no reason provided 3 06.8%
Involuntary Turnover 14 31.8%
 Program closure/Lay-off 7 15.9%
 Terminated 6 13.6%
 Deceased 1 02.3%
Missing data 3 06.8%
Open in a new tab

Total N=44. Percentages may not equal 100% due to rounding.

Structured Interviews with Former Employees

Interview data from 66 former counselors (27% of the turnover sample) and 14 former clinical supervisors (32% of the turnover sample) allowed us to directly compare the turnover information provided by the organizational representative to the data provided by the former employee him or herself. In 86% of the counselor cases and 85% of the clinical supervisor cases, classification of voluntary versus involuntary turnover was identical. In terms of the specific reason for turnover, agreement between the organizational data and data obtained from actual employees was around 70%. In most cases, the former employee cited one primary reason. This provides some reasonable assurance that the organizational data are accurate.

Ninety-one percent of counselors and 69% of clinical supervisors that we interviewed were re-employed. Of these individuals, 64% of the counselors were still in the field of substance abuse counseling whereas 89% of clinical supervisors were still in the field. Most of those counselors who were no longer in the field reported working into an occupation that was a better fit with their skills and interests (e.g., teaching). None of the individuals who turned over were re-employed by their former treatment organization. There were 6 counselors who are not re-employed. Four of them were not seeking employment, did not plan to seek employment, and did not want to be in the field of substance abuse treatment. The remaining 2 unemployed counselors were seeking employment and wanted to work in the field of substance abuse treatment. For supervisors, 4 were not currently employed. Of those, 3 were seeking employment and were hoping to stay in the field. The remaining supervisor was seeking employment, but not in the field of substance abuse treatment. This suggests that for both counselors and clinical supervisors, the turnover that occurred was primarily organizational not occupational.

Discussion

The current study provides a more precise estimate of actual counselor and clinical supervisor turnover using organizational data collected over a one year period. We also shed light on some of the specific reasons for turnover and provide information on whether the turnover that occurs tends to be voluntary or involuntary. Three overall conclusions can be reached. First, our findings suggest that turnover is a problem for the field of substance abuse treatment, perhaps even greater than some studies suggest. Second, the majority of turnover among both counselors and clinical supervisors appears to be voluntary. Third, in this sample the most common reasons for leaving were job-related.

The Turnover Problem in Substance Abuse Treatment

The voluntary turnover rate for counselors that was found in the present study (24.9%) is actually higher than the voluntary turnover rate of 19.6% reported by the Bureau of Labor Statistics (BLS) for the health care and social assistance industry (www.nobscot.com/survey/index.cfm, retrieved August 27, 2009). To place our findings in a broader context, the voluntary counselor turnover rate is also slightly higher than the total voluntary turnover rate for all occupations as reported by the BLS, which is reported as 23.4% for the year ending 2005 (our turnover data were collected in 2007) (www.nobscot.com/survey/index.cfm, retrieved August 27, 2009). The rate of voluntary turnover for clinical supervisors is also concerning (14.4%). Taken together, this indicates that the substance abuse treatment workforce may be quite unstable.

It is also noteworthy that most of the counselor and clinical supervisor turnover that occurred in substance abuse treatment organizations was voluntary. This is important because the negative effects of turnover are likely to be greater in the case of voluntary rather than involuntary turnover (Osterman, 1987; Price, 1977). This is because higher performing employees are more likely to leave voluntarily than low performing ones (McEvoy & Cascio, 1987). Future research should examine whether high or low performers are leaving substance abuse treatment.

Why are Counselors and Clinical Supervisors Leaving?

Our findings provide some insight into why counselors and clinical supervisors may be leaving. As shown in Table 1 and Table 2, a variety of factors appear to contribute to clinician turnover. The most commonly reported reason for both groups is for a new job or other opportunity. This suggests that some counselors and clinical supervisors may be proactively seeking better employment. Both employee groups also reported termination as a relatively common reason for turnover. This may be indicative of inadequate job preparation, which seems reasonable given the varied professional experiences of among counselors and limited training available for many clinical supervisors (Borders & Leddick, 1988; Culbreth & Borders, 1998; Leddick & Stone, 1982; Powell & Brodsky, 1993). Termination may also represent inadequate selection practices.

For counselors, other more frequently reported reasons for turnover include resignations and relocations. The category of resignations is admittedly vague. Post-hoc examination of interview data from the sub-sample of former employees indicates that in 26% of the cases where the organization indicated resigned the former employee discussed job dissatisfaction (e.g., dissatisfaction with supervision, pay), 8% were for personal/family reasons (e.g., illness, maternity), and 3% involved a poor occupational fit. Turnover due to relocation may be particularly difficult for treatment organizations to minimize since it is often driven by personal reasons, such as a career opportunity for one’s spouse, the desire to be closer to one’s family, or some other personal family-related reason (Eby, 2001; Stroh, 1999). The lack of a single national credentialing body make out-of-state relocation particularly complicated (Gallon et al. 2003). This may contribute to turnover not just from organizations, but also from the profession; a growing concern given workforce demand projections and our data indicating among the sub-sample of counselors that turned over, 36% ended up leaving the profession completely.

Implications for Practice

Several practical implications emerge from the present study. In light of our findings indicating that turnover in the substance abuse treatment field is perhaps even greater than previously thought, it is especially important to note that any amount of turnover is disruptive to business operations and potentially costly. Moreover, healthcare is the largest industry and seven of the 20 fastest growing jobs are health-care related (http://www.bls.gov/oco/cg/cgs035-htm, retrieved August 27, 2009). This means that not only is the demand for jobs in this industry segment substantial, it is also growing. Specifically, the projected growth for the occupation of counselor within the healthcare sector is expected to increase 29% from 2006 to 2016 (http://www.bls.gov/oco/cg/cgs035-htm, retrieved August 27, 2009) translating into a need for 45,000 new entrants into the workforce during this same timeframe (http://online.onetcenter.org/, retrieved September 15, 2009). These trends provide a compelling case for the importance of actively engaging in strategies to try to reduce turnover, particularly among counselors.

Luckily, there is considerable research on the predictors of turnover. This research suggests that workplace interventions which enhance quality of worklife and reduce workplace stress are likely to reduce turnover (Griffeth, Hom, & Gaertner, 2000). This might include providing counselors with greater autonomy and participation in work-related decisions, providing career development opportunities (e.g., training and skill-building), enhancing leadership effectiveness (e.g., selecting and promoting clinical supervisors based on interpersonal skills in addition to clinical competencies and administrative skills), improving coworker relationships (e.g., offering teambuilding or conflict management training), and clarifying role expectations to help reduce work-related stress. Research also shows that perceptions of pay satisfaction and perceived fairness of pay allocation decisions predict employee turnover (Griffeth et al., 2000). This presents a challenge to the field of substance abuse treatment since the average pay for substance abuse counselors is around $37,000 annually, which is considerably less than many other similar healthcare occupations (http://online.onetcenter.org/, retrieved September 15, 2009).

Limitations and Directions for Future Research

Like all research, the present study has several limitations. Although obtaining turnover data from actual organizational records is an advance over existing research, it is possible that some of the reasons recorded for employee turnover do not accurately represent employee motives for leaving. For example, in some treatment centers there may be hesitancy to flag an employee as “terminated” for fear of legal challenges, and instead a mutual decision may be made that the official, recorded reason for turnover is employee resignation. Fortunately, we were able to compare data from organizational records to some of the former employees’ reports of why they left. By and large there was considerable agreement, both in terms of the classification of turnover as voluntary versus involuntary and the specific reason for turnover. It is also worth noting that reliance on former employee accounts of turnover is fraught with other problems (e.g., external attributions for job loss, social desirability). So, we recognize that neither source of data are perfect and there is no way to obtain a “true score” on voluntary versus involuntary turnover, or the reason for leaving. Nonetheless, we did ask a series of questions to organizational representatives regarding the perceived accuracy of the data collected. Using a scale of 1 (not at all confident) to 10 (totally confident), the average response to a question asking about the classification of voluntary vs. involuntary turnover for each departed employee was 9.7, with 88% of respondents reporting total confidence. In terms of the specific reason given for turnover on the turnover form the average level of confidence was 9.1 and 76% reported total confidence. Taken together, this provides some assurance that our turnover data is valid.

Another potential limitation is our focus on primarily treatment centers that are affiliated with the CTN. However, when we compared the characteristics of our sample to data from counselors within two nationally representative samples of treatment facilities (as reported by Ducharme et al., 2008) we found few differences in gender (64% in our sample compared to 61% in Ducharme et al.), average age (43 years in our sample compared to 45 years in Ducharme et al.), education level (54% Masters degree or higher in our sample compared to 43% in Ducharme et al.), or certification (50% in our sample compared to 56% in Ducharme et al.). Therefore, while we cannot be sure that our data are generalizable to all treatment programs in the U.S. these comparisons suggest that our findings may be applicable to a wide range of treatment organizations. Finally, our exclusive focus on full-time employees may be viewed as a limitation since our findings are not applicable to part-time or contract employees.

In closing, the current study provides the first published, systematic perspective on the problem of turnover in the field of substance abuse treatment using longitudinal organizational data. Our findings suggest that turnover is indeed a concern for the field, especially in terms of retaining counselors. We hope that this research can serve as a springboard for additional studies of turnover among individuals in the substance abuse profession.

Footnotes

1

This project described was supported by Award Number R01DA019460 from the National Institutes on Drug Abuse awarded to Lillian T. Eby. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Drug Abuse or the National Institutes of Health. Thanks to Lisa Baranik, Marcus Butts, Lori Ducharme, Sarah Evans, Carrie Hurst, Hannah Knudsen, Aaron Johnson, Charles Lance, and Carrie Owen for their assistance with data collection and analysis.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Lillian T. Eby, Email: leby@uga.edu, The University of Georgia, 228 Psychology Building, Athens, GA 30602, 706.542.0378

Hannah Burk, Email: burk.hannah@gmail.com, The University of Georgia, 228 Psychology Building, Athens, GA 30602, 706.542.2174.

Charleen P. Maher, Email: charleen.maher@gmail.com, The University of Georgia, 228 Psychology Building, Athens, GA 30602, 706.542.2174

References

  1. Alexander JA, Bloom JR, Nuchols BA. Nursing turnover and hospital efficiency: An organizational-level analysis. Industrial Relations. 1994;33:505–520. [Google Scholar]
  2. Borders LD, Leddick GR. A nationwide survey of supervisor training. Counselor Education and Supervision. 1988;27(3):271–283. [Google Scholar]
  3. Bureau of Labor Statistics. http://www.bls.gov/oco/cg/cgs035-htm (retrieved August 27, 2009)
  4. Carise D, Gurel O, McLellan AT, Dugosh K, Kendig C. Getting patients the services they need using a computer-assisted system for patient assessment and referral – CASPAR. Drug and Alcohol Dependence. 2005;80:177–189. doi: 10.1016/j.drugalcdep.2005.03.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Carise D, McLellan AT, Gifford LS, Kleber HD. Developing a national addiction treatment information system: An introduction to the drug evaluation network system. Journal of Substance Abuse Treatment. 1999;17:67–77. doi: 10.1016/s0740-5472(98)00047-6. [DOI] [PubMed] [Google Scholar]
  6. Carroll KM, Rounsaville BJ. A vision of the next generation of behavioral therapies research in the addictions. Addiction. 2007;102:850–862. doi: 10.1111/j.1360-0443.2007.01798.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Center for Substance Abuse Treatment. Treatment improvement protocols (TIPs) evaluation project: Retrospective study wave I final report. Rockville, MD: U.S. Department of Health and Human Services; 2000. [Google Scholar]
  8. Connors GJ, Carroll KM, DiClemente CC, Longabaugh R, Donovan DM. The therapeutic alliance and its relationship to alcoholism treatment participation and outcome. Journal of Consulting and Clinical Psychology. 1997;65:588–598. doi: 10.1037//0022-006x.65.4.588. [DOI] [PubMed] [Google Scholar]
  9. Culbreth JR, Borders LD. Perceptions of the supervisory relationship. Journal of Substance Abuse Treatment. 1998;15:345–352. doi: 10.1016/s0740-5472(97)00254-7. [DOI] [PubMed] [Google Scholar]
  10. Ducharme LJ, Knudsen HK, Roman PM. Emotional exhaustion and turnover intention in human service occupations: The protective role of coworker support. Sociological Spectrum. 2008;28:81–104. [Google Scholar]
  11. Eby LT. The boundaryless career experiences of mobile spouse in dual-earner marriages. Group and Organization Management. 2001;26:343–368. [Google Scholar]
  12. Gallon SL, Gabriel RM, Knudsen JRW. The toughest job you’ll ever love: A Pacific Northwest treatment workforce survey. Journal of Substance Abuse Treatment. 2003;24:183–196. doi: 10.1016/s0740-5472(03)00032-1. [DOI] [PubMed] [Google Scholar]
  13. Griffeth RW, Hom PW, Gaertner S. A meta-analysis of antecedents and correlates of employee turnover: Update, moderator tests, and research implications for the next millennium. Journal of Management. 2000;26:463–488. [Google Scholar]
  14. Hycner RH. Some guidelines for the phenomenological analysis of interview data. Human Studies. 1985;8:279–303. [Google Scholar]
  15. Johnson JA, Roman PM. Predicting closure of private substance abuse treatment facilities. Journal of Behavioral Health Sciences & Research. 2002;29:115–125. doi: 10.1007/BF02287698. [DOI] [PubMed] [Google Scholar]
  16. Kacmar KM, Andrews MC, Van Rooy DL, Steilberg RC, Cerrone S. Sure everyone can be replaced…but at what cost? Turnover as a predictor of unit-level performance. Academy of Management Journal. 2006;49:133–144. [Google Scholar]
  17. Knight DK, Broome KM, Simpson DD, Flynn PM. Program structure and counselor-client contact in outpatient substance abuse treatment. Health Research and Educational Trust. 2007;43:616–634. doi: 10.1111/j.1475-6773.2007.00778.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Knudsen HK, Ducharme LJ, Roman PM. Research participation and turnover intention: An exploratory analysis of substance abuse counselors. Journal of Substance Abuse Treatment. 2007;33:211–217. doi: 10.1016/j.jsat.2006.12.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Knudsen HK, Johnson JA, Roman PM. Retaining counseling staff at substance abuse treatment centers: effects of management practices. Journal of Substance Abuse Treatment. 2003;24:129–135. doi: 10.1016/s0740-5472(02)00357-4. [DOI] [PubMed] [Google Scholar]
  20. Lamb S, Greenlick MR, McCarty D. Bridging the gap between practice and research: Forging partnerships with community-based drug and alcohol treatment. Washington, DC, US: National Academy Press; 1998. [PubMed] [Google Scholar]
  21. Leddick GR, Stone SC. A nationwide survey of supervision practices: A preliminary report. Presentation at the annual convention of the American Personnel and Guidance Association; Detroit, MI. 1982. Apr, [Google Scholar]
  22. Lum L, Kervin J, Clark K, Reid F, Sirola W. Examining nursing turnover intent: Job satisfaction, pay satisfaction or organizational commitment? Journal of Organizational Behavior. 1998;19:305–320. [Google Scholar]
  23. McCaul ME, Svikis DS. Improving client compliance in outpatient treatment: Counselor-targeted interventions. NIDA Research Monograph. 1991;106:204–215. [PubMed] [Google Scholar]
  24. McEvoy GM, Cascio WF. Do good or poor performers leave? A meta-analysis of the relationship between performance and turnover. Academy of Management Journal. 1987;30:744–762. [Google Scholar]
  25. McLellan AT, Carise D, Kleber HD. Can the addiction treatment infrastructure support the public’s demand for quality care? Journal of Substance Abuse Treatment. 2003;25:227–121. [PubMed] [Google Scholar]
  26. Miles MB, Huberman AM. Qualitative data analysis: An expanded sourcebook. 2. Thousand Oaks, CA: Sage; 1994. [Google Scholar]
  27. National Institute on Drug Abuse. http://www.drugabuse.gov/CTN (retrieved March 15, 2010)
  28. National Survey of Substance Abuse Treatment Services. Data on substance abuse treatment facilities. Washington, DC: Department of Health and Human Services, Substance Abuse and Mental Health Services Administration; 2007. [Google Scholar]
  29. Nobscott Corporation. www.nobscot.com/survey/index.cfm (retrieved August 27, 2009)
  30. Osterman P. Turnover, employment security, and the performance of the firm. In: Kleiner M, editor. Human resources and the performance of the firm. Madison, WI: Industrial Relations Research Association; 1987. pp. 275–317. [Google Scholar]
  31. Price JL. The study of turnover. Ames, IA: Iowa State University Press; 1977. [Google Scholar]
  32. Powell DJ, Brodsky A. Clinical supervision in alcohol and drug abuse counseling: Principles, models, methods. San Francisco, CA: Jossey-Bass; 1993. [Google Scholar]
  33. Powell MJ, York RO. Turnover in county public welfare agencies. Journal of Applied Social Sciences. 1992;16:111–127. [Google Scholar]
  34. Saxon AJ, McCarty D. Challenges in the adoption of new pharmacotherapeutics for addiction to alcohol and other drugs. Pharmacology & Therapeutics. 2005;108:119–128. doi: 10.1016/j.pharmthera.2005.06.014. [DOI] [PubMed] [Google Scholar]
  35. Staw BM. The consequences of turnover. Journal of Occupational Behaviour. 1980;1:253–273. [Google Scholar]
  36. Stroh LK. Does relocation still benefit corporation and employees? An overview of the literature. Human Resource Management Review. 1999;9:29–308. [Google Scholar]
  37. United States Department of Labor. http://online.onetcenter.org/ (retrieved September 15, 2009)
  38. Whitter, Bell, Gammond, Gwaltney, Magana, Moreaux Strengthening professional identity: Challenges of the addictions treatment workforce. 2006 Available from http://www.samhsa.gov/workforce/workforcereportfinal.pdf (retrieved September 15, 2009)

RESOURCES