Association for Public Policy Analysis & Management

Nearly half of child care workers in the United States receive some sort of public assistance. Their wages, on average, are less than 10 dollars an hour. A major National Research Council report entitled “Transforming the Workforce” recently called out these patterns and argued that addressing the poor working conditions many early childhood educators face is essential for ensuring young children have stable, qualified teachers.

Although policy interest in the early childhood education (ECE) workforce has grown rapidly, rigorous research in this area has been limited, in large part due to data limitations. The rich longitudinal data systems available to track teachers and students in the K-12 system have allowed for extensive research on that labor force. In contrast, in the fragmented ECE market, which includes private providers, Head Start centers, and state-run preschools, system-wide data collection does not occur. This lack of data has severely limited our understanding of this labor force.

No previous studies have tracked a large sample of ECE teachers longitudinally, and none have examined the relationship between teacher turnover and program quality. This study provides the first empirical evidence on system-wide ECE teacher turnover.

In 2012 the Louisiana Department of Education introduced a Quality Rating and Improvement System (QRIS) that was unique in two ways. First, the QRIS is mandatory for all publicly-funded programs, including state pre-kindergarten, Head Start, and child care centers receiving public subsidies. Second, the QRIS is focused on the quality of teacher-child interactions, as measured by the Classroom Assessment Scoring System (CLASS), a widely-used measure. Since 2015, Louisiana has collected CLASS data multiple times per year in every single publicly-funded ECE classroom. Importantly, these data allow us to match individual ECE teachers over time, thus providing the first opportunity to track ECE teacher turnover across an entire system and to link these data to information about teacher-child interactions.

Our sample includes teachers in over 1,500 publicly-funded programs operating in both 2015-16 and 2016-17. We find that 36 percent of all ECE teachers observed in 2015 were no longer at their same program one year later. Turnover rates were highest in child care centers where nearly half (47%) of all teachers observed in 2015 were no longer there one year later, and lowest among preschool teachers in schools (23%).

Exiting teachers have CLASS scores that are nearly half a standard deviation lower than those of stayers. Further, teachers who exit are replaced by teachers whose CLASS ratings are ~ .125 standard deviations higher. Overall, we find that roughly half of the overall gains in CLASS scores statewide over this period are explained by increases to teachers’ CLASS ratings over time, while the other half is attributable to the higher CLASS scores of program entrants relative to exiters.

Our results highlight very high turnover rates, especially in child care settings. Although existing research indicates this turnover can be harmful to young children, our results suggest that in some cases teacher turnover may be an important driver of quality improvements.