Into—and Out of—City Schools: The Retention of Teachers Prepared for Urban Settings

Abstract

This retrospective, longitudinal study compares the career paths of graduates of Harvard University's master's-level teacher education program (TEP) before and after it adopted an urban emphasis in 2001. I surveyed all graduates of TEP, 1985–2006, which yielded a sample of 636. Using discrete-time survival analysis, I examined where respondents taught and whether, when, and why they left teaching. I found that graduates of TEP's urban-focused program were more likely to teach in urban schools than the comparison cohorts but less likely than prior cohorts to stay in teaching. However, compared to the non-urban cohort members that were working in urban schools, urban cohort members were no more likely to exit. I further found that Whites and males in the urban cohorts were more likely to exit than members of these groups in previous cohorts, whereas females of color who prepared for urban settings were much less likely to leave than their counterparts who had received more general preparation. A large number of urban cohort members who left teaching identified “lack of community support for the school” as a key factor in their decision to leave teaching. At the same time, they felt drawn into other roles, such as principal, or professions, such as law, that, in their view, enabled them to have greater impact on urban schooling.

Morgaen L. Donaldson is an assistant professor of educational leadership at the Neag School of Education, University of Connecticut. Her research focuses on teacher quality, teacher retention, school leadership, and teachers' unions, particularly in schools and districts serving low-income children and children of color.

Notes

∗ p <.05,

∗∗ p <.01.

¹Sample sizes differ slightly by cell because some respondents did not provide data on certain covariates.

∗p <.05;

∗∗p <.01,

∗∗∗p <.001; chi-square tests compared each comparison group to the urban cohorts.

∼ significant at 10%;

∗significant at 5%;

∗∗significant at 1%; standard errors in parentheses (all compared to baseline model with urbcohort (M1 and 2) or recent (M3 and 4) as single predictor).

∗significant at 5%;

∗∗significant at 1%; standard errors in parentheses

¹Pearson chi2(4) = 8.2558 Pr = 0.083 (urban vs. prior 5)

²Pearson chi2(4) = 13.6353 Pr = 0.009 (urban vs. prior 5)

1. I used five time dummies because urban-focused graduates could have taught for five years, maximum. In other analyses, I included up to 21 time dummies to estimate when TEP's pre-2001 graduates were most likely to leave.

2. 636 individuals responded to the survey. 47 of these respondents had never taught in K-12 schools. An additional 11 respondents did not answer the questions on their school demographics.

3. For example, in recent years there have been several well-publicized stand-offs between teachers unions and management in large, urban districts such as San Diego and Washington DC.