Who Exits When? a Survival Analysis of Departures from Frontline Child Welfare Work

Background: Turnover among the frontline child welfare workforce is high averaging 30%. High caseworker turnover is a central disrupter of an efficient, effective child welfare system. Studies indicate that most workers leave their positions or agencies within two years of hire. However, “departure” when broadly defined, does not account for promotions or lateral moves within child welfare. Little is known about workers’ exits from the field. Drawing from the theory of embeddedness which highlights the importance of connections in workers’ employment decisions, this study builds upon current evidence by considering three central questions: (1) when do child welfare workers exit the field? (2) what personal and support characteristics shape employment trajectories? and (3) do the influence of these characteristics on turnover change over time?

Methodology: This study uses data from The Florida Study of Professionals for Safe Families, a longitudinal cohort study of newly-hired workers in frontline casework positions. Participants were recruited during pre-service training classes and followed for 3.5 years (N = 1,500). Data were collected electronically at 6-month intervals to assess individual, organizational, and occupational attributes and wellbeing. Response rates across waves averaged 81%. We used Cox survival regression models to examine whether and when departures from child welfare occurred in the first 3.5 years after hire. The time-to-event analyses considered how the variation in the timing of exits varied by worker’s demographic characteristics and workplace support at the time of hire (e.g., work-family fit, Spector Job Satisfaction Survey domains) and how the influence of these predictors changed over time.

Findings: The survival function indicated that workers departed the field quickly such that only 31.5% were left in child welfare at 3.5 years. The median lifetime in child welfare among this cohort was 24.2 months, just over 2 years. The time of greatest risk was around 12 months with smaller peaks at 24 and 38 months suggesting workers depart at anniversaries, approximately 1, 2, and 3 years after hire. Regression results suggested that individual and support characteristics were largely related to departure times. Respondents with higher levels of education or those with social work degrees exited earlier than their counterparts without such degrees. Workers who felt that the job could interfere with relationships or children also averaged shorter times to departure. Perceiving supervisory support promoted staying in child welfare as did satisfaction with pay and promotion. Although the importance of at-hire supervisory support declined over time, the importance of satisfaction with pay and promotion did not.

Conclusions and Implications: Understanding the length of time to departure, as well as factors that affirm or constrain employment decisions, can assist agency administrators in developing on-boarding and retention practices. Increased risk of departure at work anniversaries suggest opportunities for organizational strategies to support workers’ career development at specific times in their employment trajectories. Related, agency leaders may benefit from identifying organizational strategies to attenuate fit challenges for workers and consider their professional development and opportunities for growth within the agency as ways to promote retention in the field of child welfare.