Association for Public Policy Analysis & Management

Children’s experiences in early childhood education (ECE) settings can have substantial impacts on their skills at school entry, and these skills have long-run implications for educational, labor market, and wellbeing outcomes. Research increasingly suggests that the adults who care for children in these settings are integral to building children’s early skills. For this reason there has been growing public investment in professionalizing and improving the ECE workforce, which has long been characterized by low wages, low levels of education, and high turnover. One primary goal of these recent efforts is to “stabilize” the ECE workforce by reducing turnover which, it is theorized, will lead to improved child outcomes. Surprisingly, however, there is little empirical research aimed at understanding the drivers of ECE teacher turnover and next to none examining the link between teacher turnover and child outcomes. This study fills this gap by examining within-year teacher turnover in Head Start, a large, federally-funded ECE program serving the United States’ most vulnerable children.

We use the 2006 and 2009 waves of the nationally-representative Head Start Family and Child Experiences Survey (FACES, N≈4,000 children, N≈850 teachers, N≈240 centers). We begin by examining the relationship between within-year teacher exits—that is, teachers who leave during the Head Start program year—and an extensive set of center-level factors such as compensation and benefits, director characteristics, opportunities for professional development, and teacher and director measures of center climate. We find that 10% of children in Head Start have a teacher that leaves Head Start entirely during the program year. This within-year turnover is systematically related to center climate and leadership. For example, after accounting for an extensive set of teacher characteristics, we found that turnover was strongly predicted by aggregated teacher reports of center support; that is, their sense that Head Start provided assistance in the classroom, appropriate help for new teachers, and support for teamwork. Turnover was also associated with inexperienced directors.

The key goal of our analysis was to examine the relationship between teacher turnover and child outcomes. To do this, we use a fixed-effects analysis to measure the effects of turnover on children’s development across a broad set of outcomes including measures of mathematics, literacy, and behavioral regulation. Our results suggest that within-year teacher turnover is deleterious for children’s development, particularly for mathematics and behavioral outcomes. Effect sizes were modest. For example, within-centers, teacher turnover was associated with a -0.14 of a standard deviation reduction in children’s growth in mathematics.

Our findings provide the first empirical evidence on within-year ECE teacher turnover, using large, nationally-representative data. The results are consistent with recent claims that within-year turnover has negative impacts on children’s development, and suggest that policies aimed at improving center supports for teachers may offer an important avenue for reducing turnover.