Purpose: Little is known about professional impairment, or the degree to which social workers continue to work when they are too distressed to practice effectively. Most impairment research is based on samples of other professionals, and most published articles are conceptual or descriptive. The purpose of this study is to investigate 1) a model wherein burnout mediates the relationship between work stressors and impairment and 2) a model wherein personality and work support factors mediate the relationship between work stressors and burnout in model 1. Based on previous research, we also hypothesized that income (having enough income to cover basic needs) moderates burnout.
Methods: The study investigated personal and occupational issues of actively practicing NASW members by anonymously surveying a probability sample in North Carolina. Three-hundred-eighty-two respondents completed a mailed questionnaire that included measures of work stress (e.g., facing ethical compromises, experiencing discrimination), workplace supports (e.g., supportive supervision, coworker support, material resources), personality (e.g., needing approval, having difficulty asking for help), burnout (i.e., Maslach burnout subscale), and impairment (i.e., working when too distressed to be effective). Summary variables were created by factor analyzing like variables using exploratory factor analysis (i.e., principal axis factoring with an oblique rotation). All data met the analytic assumptions, and correlations, one-way ANOVAs, and SEM analyses were conducted.
Results: The sample was demographically comparable to the national population of NASW members. As expected, workload variables were not related to either burnout or impairment in bivariate analyses, so were not included in the analysis. In model 1, the relationship between work stressors and impairment decreased when burnout was added, revealing mediation. In both models, work stressors and burnout variables directly influenced impairment (.042 and .064, respectively). When the additional variables were added in model 2, work stressors direct influence on burnout decreased by 25 percent and the measured disturbance (error variance) decreased. All relationships in Model 2 were significant except for the effect of work stressors on personality, so the model was changed to designate work stressors as a burnout moderator. This final model had excellent fit [X2(5, N = 382) = 5.352, p =0.3744; RMSEA = .014; SRMR = .021; CFI = .999; TLI = .997]. All relationships were significant (p < .03), and 32.6% of impairment variance was explained.
Implications: Because 166 (44%) of the social workers agreed or strongly agreed that they worked when too distressed to be effective, this is a serious issue for these practitioners’ clients and supervisors. Burnout is more than a personal issue for social workers; it is related to diminished quality of client care. The model in this analysis suggests several intervention opportunities. For example, improving workplace supports could buffer the influence of work stressors on burnout. Similarly, providing sufficient income could reduce burnout, as could mentorship around issues like asking for help and needing approval. Rather than urging practitioners to practice “self-care” strategies (unrelated to burnout in these analyses), supervisors should find ways to reduce stressors (e.g., harassment, discrimination) and support their practitioners in the workplace.