Groundhog Day or Musical Chairs? The Persistence of School Staff Vacancies

• In Event: Filling Teacher Vacancies

Abstract

Policies aimed at alleviating school staffing shortages frequently target specific schools or districts. Federal teacher loan forgiveness programs, for example, direct benefits to schools serving low-income students while state education agencies are increasingly identifying high-shortage districts for additional funding or technical assistance. These approaches rest upon an implicit assumption that the schools and districts experiencing shortages today will continue to experience them in the future. While intuitive, if shortages are largely transitory and driven by idiosyncratic events, such as individual retirements or position changes, then targeting based on current conditions may misidentify future need. Despite the importance of this assumption for policy design, extant research offers little empirical evidence on the extent to which school staffing shortages persist over time. 

I examine the temporal stability of school staffing vacancies using administrative data from Illinois public schools. Since 2017, the Illinois State Board of Education has collected school-level data on unfilled positions as part of its Employment Information System. These data offer several advantages over other shortage proxies. Unlike measures based on certification status or preparation program enrollment, unfilled positions provide a relatively direct indicator that a school is unable to provide intended services to students. Unlike survey-based measures, these data are part of regular data collections, reducing susceptibility to nonresponse and recall bias. And unlike aggregate supply-demand models, they capture variation at the school level which is especially important given that shortages vary both between and within districts (Bruno, 2025; Edwards & Kraft, 2024). I use data from the 2022-23 through 2025-26 school years, a period during which reporting requirements were fully established and vacancy data were consistently reported at the school level. 

Preliminary results suggest that school staff vacancies are moderately persistent for teaching positions but substantially transitory overall. A variance decomposition of teacher vacancy rates among schools experiencing any vacancies indicates that approximately 42% of variation is attributable to school characteristics while 58% reflects year-to-year fluctuations. For administrator vacancies, between-school persistence is essentially zero, though at the district-level 37% of variance is stable, suggesting that chronic administrative hiring difficulties manifest across different schools within the same district over time. 

Rank-rank correlations of school-level teacher vacancy rates decline from approximately 0.48 between adjacent years to 0.37 at a three-year lag, indicating moderate but decaying predictive value. A simulated targeting exercise finds that schools in the top decile of teacher vacancy rates are roughly four times more likely than chance to remain there three years later, but more than 60% of initially targeted schools exit the top decile within that window.These preliminary findings indicate that most of the variation in school staff vacancies is not explained by stable school or district characteristics. These results have implications for policies that assume current shortage patterns will persist long enough for targeted interventions to reach the right schools. Future analyses will examine whether time-varying school characteristics improve prediction, whether these patterns differ across specific position types, and the extent to which these results may be driven by existing policies in the state.

Author

• Andrew Mitchell Camp, Brown University
  Presenting Author