Association for Public Policy Analysis & Management

Over the past decade, research on the effect of teachers on student performance has spurred increased attention on teacher selection, assessment, and retention. Many schools struggle to hire qualified teachers, particularly in the fields of math, science, special education, and bilingual education, with high-poverty, high-minority, and urban schools being especially hard to staff. While the supply and recruitment of teachers are important to fill the demand for new hires, evidence suggests that teacher retention is especially challenging for schools. In New York City (NYC), for example, about a quarter of middle school teachers leave within a year of entering the workforce and more than half within the first three years. While some level of teacher mobility can be beneficial for schools, high turnover increases the number of inexperienced teachers, reduces teacher quality, disturbs school-community relationships, and costs time and money. In this paper, we analyze the impact of a long-standing middle school science initiative in NYC on teacher retention.

Urban Advantage (UA) is a collaboration between eight science-rich cultural institutions (museums, zoos, and gardens) and the NYC Department of Education (NYCDOE), which began in 2004. It is designed to provide teachers and students in middle school (grades 6-8) the opportunity to engage in authentic science practice through classroom materials, professional development, and access to cultural institutions. Recent evidence shows that the program has successfully improved students’ science achievement. Roughly half of NYC middle schools have at some point participated in the program since it began in 2005. In 2015-16, over 500 teachers across the district participated. UA has also attracted the attention of school districts and institutions in other cities (e.g. Denver, Kansas City, and Boston) and countries (e.g. Israel and the United Kingdom) interested in providing similar programs.

The goal of this paper is to test the hypothesis that, all else equal, science teachers who participate in the UA program are less likely to leave their school or the district compared to their non-UA counterparts. To do so, we exploit detailed teacher-level administrative personnel data from the NYCDOE and program participation data on individual teachers participating in UA. We use a discrete-time hazard model that accounts for teacher and school unobserved heterogeneity to adjust for selection into the program, and a rich set of observable teacher and school characteristics. Preliminary findings suggest that UA science teachers are close to 6 percentage points less likely than non-UA science teachers to leave their school in the following academic year, and about 3 percentage points less likely to leave the district.

Urban school districts have a unique opportunity to take advantage of the concentration of science-rich cultural institutions around the city. Results from this paper will contribute to the growing literature on teacher retention, and, more importantly, innovative and less costly measures school districts can take to improve teacher retention. Given evidence that urban districts are especially burdened with high teacher turnover, NYC, the nation’s largest (urban) school district, is an excellent setting to learn best practices for teacher retention.