Abstract
Counselor turnover is a serious problem for substance use disorder (SUD) treatment programs, especially in rural areas. This study examined the relationship between workload, workplace environment, and emotional exhaustion on turnover intentions among SUD treatment counselors in rural Pennsylvania. An online, anonymous survey was administered to rural SUD treatment counselors in Pennsylvania between October and December 2020. Completed surveys from 206 counselors were used for analysis. Variables included multi-dimensional measures of emotional exhaustion, intention to quit, workload, and workplace environment. Results showed age, perceptions of distributive justice, and management communication to be inversely related to emotional exhaustion and intention to quit. A subsequent analysis showed that the direct effects of these variables on intention to quit were reduced with the inclusion of emotional exhaustion in modeling, suggesting a potential mediation effect. Findings indicated that higher levels of emotional exhaustion were directly related to a greater intention to quit. A statistically significant indirect relationship was also found for distributive justice on intention to quit through emotional exhaustion, indicating a significant mediating effect. While counseling is a stressful occupation with high potential for burnout, these results show that organizations can adopt practices that help lower emotional exhaustion and turnover intentions for their counselors which could improve retention in the SUD treatment field. This is especially important for rural areas, where there is a high demand for SUD services and not enough treatment workers to meet that demand.
Highlights
Rural SUD counselors experience high levels of burnout and turnover.
Emotional exhaustion (“burnout”) is the strongest predictor of intent to quit.
Management communication and distributive justice can improve counselor retention.
Keywords:
Notes
1 The Center for Rural Pennsylvania defines a municipality as rural “when the population density within the municipality is less than the statewide average density of 284 persons per square mile, or the total population is less than 2,500, unless more than 50 percent of the population lives in an urbanized area as defined by the U.S. Census Bureau” (https://www.rural.palegislature.us/demographics_rural_urban.html).
2 Only 14 respondents who provided complete data indicated a racial identification other than White.
3 The number of bootstrap samples was set to “2000” and bias-corrected confidence level set to “95.” Variables were allowed to correlate with each other in modeling.
4 The term “predictors” is used to refer to a linear relationship, not a causal one.
5 The Chi-square statistic is sensitive to sample size (West et al., Citation2012). Better estimates of model fit are the model fit indices provided.