Evaluating the Effect of New-teacher Induction Programs on Teacher Turnover

New-teacher induction programs are widely used as non-monetary interventions, with the reduction of teacher turnover being one of their purposes. During the last three decades, states have been active in legislation to mandate induction programs for new teachers in public schools. Motivated by the discrepancy of the estimated program effects on teacher turnover in the existing literature, this study attempts to examine whether analytical methods used to address the endogeneity of induction-program participation matter in explaining this inconsistency in the estimation of the effect of new-teacher induction programs and how mandatory induction legislation and/or policy are linked with new-teacher induction programs and teacher turnover. Specifically, three key research questions are examined, including what are the determinants of new teachers' turnover, whether new-teacher induction programs have effects on teacher turnover, and whether state induction legislation and/or local mentoring policy have effects on teacher turnover. Two main sources of data are employed, including multiple cycles of Schools and Staffing Survey and Teacher Follow-up Survey, the largest nationally representative sample surveys on teachers and their schools in the United States, as well as New York City Department of Education administrative data, the longitudinal data concerning public school new teachers and their job status in the largest public education system in the United States. In order to address the potential endogeneity problem that may be associated with the induction program participation, this study employs identification strategies such as propensity-score matching, instrumental variable, and difference-in-difference methods. Particularly, in estimation, the difference-in-differences strategy is incorporated with the instrumental variable method as well as nonlinear models. The current study finds that, after addressing the endogeneity problem, new-teacher induction programs hardly have any effects on reducing teacher turnover, which is different from what most previous studies have found. It is also implied that if a mandatory induction/mentoring law/policy does not encourage new teachers' participation in a comprehensive induction/mentoring program, it may not work in terms of reducing teachers' exit turnover at best. This study contributes in literature on new-teacher induction programs and teacher turnover in terms of analytical methods, generalizability of findings, understanding of induction legislation and policy, and proper research design considerations for future studies.