State pilot finds high turnover in child care workforce; pilot pay boosts showed limited retention effects and produced some unintended behavior

Dr. Barnes of the Department of Education presented findings from a two‑year randomized pilot that tested pay supplements to early childhood teachers.

The pilot randomly assigned participating centers to one of three groups: a fixed dollar supplement (about $2,000 per teacher annually in the pilot design), a 10% pay increase for teachers at participating sites, and a control group that received no supplemental pay through the pilot. Centers were asked to list all teachers on the initial roster and the study tracked retention and replacements across eight surveys over two years. Dr. Barnes described the approach as a randomized control design intended to identify retention differences while balancing center characteristics across groups.

Key quantitative findings presented: the study observed an average turnover rate in the range of roughly 45–50% over the study period (Dr. Barnes cited prior analyses that place a one‑year turnover nearer 46% and two‑year turnover in the 50–60% range; a national rough benchmark of 30–40% was used as context). The pilot included thousands of teachers and centers (the presenters said the dataset covered across-the-state sites and that ‘‘this is over 7,000 teachers’’ in one referenced figure).

The statistical analysis did not find a clear, statistically significant difference in retention between the sites receiving a flat stipend and those receiving a 10% boost. Presenters noted several complicating factors: inconsistent data collection at center level (the department relied on center directors as the point of contact rather than surveying every teacher directly), the study’s timing soon after the COVID workforce disruptions, and local labor market changes (competing nontraditional employers such as delivery and retail that offered higher immediate pay). Presenters described additional predictive modeling that suggested substantially larger pay increases would be required to lower turnover to levels typically viewed as acceptable (for example, modeled pay near roughly $15–$16 per hour or higher was associated with much lower predicted turnover in the model used).

Qualitative and operational findings emerged during the Q&A and from center feedback. Commenting center directors reported that some teachers declined supplements because the extra pay could affect their eligibility for public benefits (Medicaid, SNAP), while others used supplemental payments as short-term cushions and called in to work more frequently, producing operational strain for centers. Several directors said offering free or prioritized childcare to teachers reduced turnover more than small pay supplements; one director said centers that prioritized teacher childcare saw turnover drop to very low levels. Presenters and participants also noted that higher center quality, higher baseline pay and teachers’ higher formal education correlated with lower turnover.

The department’s team acknowledged dataset limitations (missing or inconsistent inputs from some centers and the lack of a planned qualitative component) and said they plan additional analysis and will return to the advisory council with updates at a future meeting. The commission invited staff to refine the research design for future rounds, add qualitative fieldwork, and consider benefit effects and targeted strategies for teachers returning from childbirth or other life events.