The Role of Organizational and Work Group Characteristics On Child Welfare Workers' Turnover Intentions

Background and Purpose: Research on personal and organizational factors related to retention of child welfare workers has become an area of greater interest in the last decade with researchers increasingly recognizing the need to use multivariate analyses (DePanfilis & Zlotnik, 2008). However, thus far conceptual and analytical retention models have not clearly distinguished between organizational and work group influences. Informed by health services research findings on the importance of team psychological safety to group member learning and performance (Edmondson, 1999), engagement in quality improvement work (Nembhard & Edmondson, 2006), problem-solving efficacy and system improvement (Tucker, Nembhard, & Edmondson, 2007) and care provider engagement (Rathert et al, 2009) we hypothesized that team psychological safety, defined as shared beliefs among work unit members that the group will not embarrass, reject or punish someone for speaking up, would be related to worker retention.

Methods: Separate surveys were administered statewide to public child welfare front line social workers and supervisors with an overall response rate for social workers of 96% (n=1479) and 69% (n=164) for supervisors. Supervisory surveys were collected through a combination of mailed, paper-based forms while the majority of social worker surveys were collected via Survey Monkey, a web-based survey tool. Psychological safety was measured using the original seven item scale developed by Edmondson (1999) with a five point response Likert scale. Turnover intention was measured using a combination of the normative commitment and short-term turnover intention scales used by Moynihan and Pandey (2007). Cronbach's alpha for psychological safety was .81 and .75 for turnover intention.

Results: The initial results of structural equation modeling indicate that, while controlling for supervisory unit-level clustering, a worker's perception of supervisor's support and an office's overall focus on human resources are significantly and positively associated with a worker's level of team psychological safety. In turn, psychological safety is significantly and positively associated with a worker's desire to remain employed in public child welfare. Additionally, our analysis indicates that worker psychological safety is lower for minority workers than for White workers.

Conclusions and Implications: High turnover rates and high levels of turnover intention have characterized social work in child welfare settings for several decades (e.g. Harrison, 1980; Annie E. Casey Foundation, 2003). The current study contributes to the broader literature addressing child welfare worker turnover intention by assessing the hypothesized association between turnover intention and work-unit psychological safety. Also, this study appears to be the first study of child welfare staff retention that adequately accommodates supervisory unit-level clustering of variables. In addition to confirming the hypothesized relationships, our analysis suggests public child welfare agencies can contribute to work-unit psychological safety through supervisory training that develops the knowledge and skills needed to create and sustain work group structures and processes that support learning and task performance, as well as by directing attention to office level human resource policies and practices.