The Promise of the Residency Model: A Statewide Examination of Effects on Teacher Employment, Performance, and Retention

• In Event: Teacher Recruitment, Preparation, and Workforce Diversity

Abstract

The residency model - in which teacher education providers (TEPs) partner with school districts to provide context-specific preparation - typically leverages targeted recruitment, yearlong placements, financial support, and sustained induction of candidates in order to directly address teacher shortages in districts with persistently high rates of turnover, lower-income communities, and hard-to-staff subject areas (Berry, 2008; Guha et al., 2016). Furthermore, many residencies seek to diversify the workforce by, for example, explicitly working to attract more teachers of color (NCTR, 2016).

Residencies have risen rapidly in prevalence over the past decade. In some ways, this growth appears warranted, given promising trends of the model’s impacts on graduates (Guha et al., 2016). Existing literature suggests that residencies have found success recruiting and retaining teachers of color and teachers working in shortage subject areas (Matsko et al., 2020; NCTR, 2016; Perlstein et al., 2014). Evidence of their effects on instructional quality, though, is mixed, with graduates performing at times better, comparably, and worse than other new teachers in value-added to student achievement and feelings of preparedness to teach (Sloan & Blazevski, 2015; Papay et al., 2012).

However, as residencies are relatively new to the teacher preparation landscape, it should not be surprising that their evidence base has limitations. First, the vast majority of research relies on descriptive methods, making it impossible to conclude that the teacher preparation experience (as opposed to pre-existing differences) is the cause of any contrast, much less disentangle whether features of the residency model itself or other unrelated programmatic aspects are responsible. Additionally, most existing research comes from non-peer reviewed working papers and reports focused only on individual programs and/or published by organizations invested in the model’s success. Finally, few studies consider how outcomes vary for different groups of teachers, especially those targeted by residencies (e.g., STEM teachers). As a result, more research is needed that (1) uses causal inference methods, (2) takes a labor market perspective across many residencies simultaneously, and (3) considers how average effects differ across subgroups of graduates and subject areas.

Our study does all three. Through surveys, we have collected extensive information for almost all TEPs in STATE, including the extent to which they consider themselves residencies, the specific “residency features” (e.g., yearlong clinical placement) they have implemented, and longitudinal information about when the model and each feature were adopted. Combining this with statewide administrative datasets on program completion, personnel, and teacher evaluation, we use a generalized difference-in-differences approach to obtain credibly causal estimates of the effects of adopting the residency model (and each of its features) on graduates’ employment, retention, and instructional effectiveness. Furthermore, we explore whether these effects vary as a function of graduate racial and ethnic identity or subject area concentration. Initial results align with prior literature, suggesting that adoption of the residency model increases graduates’ likelihood of finding employment and, to a lesser extent, their instructional effectiveness, particularly for graduates of color and those in high-needs subject areas, though these effects are somewhat offset by increased mobility and attrition.

Authors

• Matthew Truwit, University of Michigan

• Emanuele Bardelli, University of Michigan - Ann Arbor

• Matthew Ronfeldt, University of Michigan