Purpose: Many child welfare agencies across the country are experiencing a severe workforce crisis involving high staff turnover rates. In 2001, a research team representing university-state partnership embarked on a workforce retention study to identify factors that influence turnover and retention in public child welfare systems. This paper presents findings from research carried out in 24 district agencies. The purpose of this study was to analyze the commonalties, similarities, and differences among 12 high turnover districts (HTD) and 12 low turnover districts (LTD) with an explicit focus on organizational and supervisory factors. The findings contribute to the emergent knowledge base on workforce turnover and retention.

Methods: Six hundred and fifty workers and supervisors from twenty-four child welfare agencies participated in a survey addressing supervisory and organizational factors related to workforce recruitment, retention, and turnover. The dependent variable was intention to leave. Independent samples t-tests were performed to assess the mean difference between the two cohorts on organizational factors such as: clarity and coherence of practice, Work-life fit, Efficacy and job satisfaction, Job supports and commitment, Technology and training, and Salary and benefits. Supervisory factors were also compared. Logistic regression models were conducted to help answer three questions. First, what effect do the organizational factors have on the likelihood of being from one of three low turnover outlier districts? Second, among the HTDs, what effect do the organizational and supervisory factors have on the likelihood of intention to leave? Third, among the LTDs, what effect do the organizational and supervisory factors have on the likelihood of intention to leave?

Results: Fifty-four percent of the respondents in the low turnover districts reported looking for another job in the past year, compared to 72% of respondents in the high turnover districts who had considered looking for another job. Mean scores on 5 of the 6 organizational factors were significantly higher in the LTD's than the HTDs. Technology and training did not differ significantly. The largest mean difference was found in satisfaction with salary and benefits (t [df, 648] =-10.74, p<.001), followed by job supports and commitment (t [df, 648] =-4.50, p<.001). Interestingly, neither supervisor support nor supervisor competence were significantly different between the two cohorts. An unexpected finding was that 72% of the respondents in three of the low turnover systems reported looking for another job. Similar to the high turnover districts, the respondents in these “outlier districts” reported less favorable perceptions of the organization climate. Logistic regression results indicated that perceptions of life-work fit significantly increased the likelihood of workers' intention to leave in both high and low turnover systems.

Implications for Practice: Comparing the differences between the low turnover and high turnover cohorts will enhance the knowledge base about the factors encourage retention and turnover. These findings may inform the development of innovative supervisory and organizational interventions aimed at enhancing retention and reducing turnover. They also pave the way for the development of empirically-grounded theoretical frameworks, which will improve future research and practice.