A multilevel analysis of organizational justice and intention to leave in social work organizations

Background & Purpose: Social workers' turnover represents practical problems to social work organizations in terms of service discontinuance and loss of skilled practitioners. The reduction of turnover rate among social workers has been a great concern and challenge for social work administrators. Previous literature has consistently demonstrated that “intention to leave” is one of the strongest predictors and an immediate precursor of social workers' turnover (Loi et al., 2006). A number of studies have found significant associations between the psychological and demographic traits of social workers, such as stress, job satisfaction, and burnout, and intention to leave (Harrington et al., 2001). Recent organizational theory, however, emphasizes the leverage effect of organizational characteristics on the relationships between intention to leave and personal characteristics (Ngo et al., 2005). In particular, current organizational research has shown increasing interest in “organizational justice” as a potential organizational factor to create benefits for organizations and employees (Cropanzano et al., 2007). Despite its importance, not much research has examined the role of organizational justice on intention to leave in social work settings. To fill this gap in research, this study empirically investigated the impact of organizational justice on the intention to leave of social workers.

Method: We used multistage cluster sampling in which we first chose 51 community welfare centers and subsequently selected an average of five social workers within these chosen centers. The total sample size finally analyzed was 218. Trained research assistants visited the centers and distributed structured questionnaires to each social worker. Organizational justice, a main independent variable, was measured by Moorman's (1991) index and the Cronbach-alpha for the index was .78. For intention to leave, a dependent variable, we modified the scale devised by Mobley (1982). The Cronbach-alpha for this scale was .82. We also measured other individual factors of social workers, such as stress, and the organizational factors of the centers, such as number of staff. By employing the “multilevel modeling” technique, this study dealt with the unique nature of hierarchical data structure, where individuals (e.g., social workers) are hierarchically nested within organizations (e.g., community welfare centers). As such, this study sophisticatedly analyzed the cross-level interaction between individual and organizational factors.

Results: The results of multilevel modeling revealed that 22.5 percent of the variation in intention to leave is explained by the differences in the organizational factors of the centers. The results also indicated that organizational justice decreases the intention to leave of social workers. Another significant finding is that organizational justice not only directly affects intention to leave, but also mediates the relationship between individual factors and intention to leave. The results of cross-level interaction analysis demonstrated that while stress, an individual factor, increases intention to leave, its effect is alleviated by organizational justice.

Implications: The findings of the study suggest that enhancing organizational justice is an efficient strategy to diminish the intention to leave of social workers in social work agencies. This study also shows how to apply multilevel modeling to hierarchical data structure, which is common in organizational research.