Comparing Predictors of 6- and 12-Month Job Exit Among Recently Hired Child Welfare Workers

Background: The effectiveness of public child welfare services has long been impacted by instability in its workforce. National turnover rates are high and frequent changes in case workers delay permanency decisions and impact other child and family outcomes. Reducing turnover in the child welfare workforce is key in protecting children from harm. In this study, we examine predictors of 6-month and 12-month departure among recently hired child welfare workers.

Methods: Data come from the first three waves of a state-wide longitudinal study investigating retention among child welfare workers.  Case managers and child protective investigators were recruited from the cohort of workers hired between September 2015 and August 2016. Wave I (W1) data were collected during pre-service training, prior to receiving cases. Wave II (W2) data were collected at 6 months, representing about three months of training and three months of actual casework.  Wave III (W3) data were collected at 12 months.

Hierarchical logistic regression procedures were used to examine the impact of early employment experiences on turnover decisions by newly hired workers.  W1 data includes personal characteristics (i.e., educational background, age, and history of childhood maltreatment), perceptions of supervisor and co-worker support, and satisfaction with pay and benefits. W2 data includess transition measures such as number of cases received at first week of independent casework and receipt of specialized mentoring. Analyses of 6-month departure use data from the entire sample (n=1,001); analyses of 12-month departure (n=605) excludes cases of workers who left their positions by W2. Cases of workers who remained in their positions across both waves served as the comparison group.

Results: At six months, about 18% (n=184) reported job exit; by 12 months, an additional 29% (n=178) had left their original positions. Personal characteristics and transition measures predicted departure at six months. Workers with social work degrees were 2.6 times more likely to leave (p=.006); those with other human service degrees were 2.1 times more likely to leave (p=.013), compared to those without any human service degree. Age was positively associated with the likelihood of 6-month departure. Transition measures also played an important role in early departure. Each additional case assigned during the first week following training resulted in an 11% greater likelihood of leaving (p=<.001), while workers receiving specialized mentoring were 35% less likely to leave early (p=.029). In contrast, the only significant early employment experience predicting 12-month departure was satisfaction with pay at W1; a one unit increase in salary satisfaction decreased the likelihood of departure by 24%.

Conclusions: These results suggest different patterns of experiences influencing employment decisions between those who depart at six months and those who depart at 12 months. Among 6-month leavers, the transition period from training into independent casework is crucial. By 12 months, however, personal characteristics and most early employment experiences no longer appear to influence departure . Building on these findings, future research should examine  more fully the differences across employment decision trajectories with attention to identifying factors of departure that are relevant as employment tenure increases.