Teacher Supply and Long-Run Student Outcomes: Evidence from World War II

• In Event: Policy impacts on the K12 Teacher Workforce

Abstract

Teacher shortages are a persistent challenge in American K–12 education, with 86% of public schools reporting difficulty hiring teachers entering the 2023–24 school year. Yet despite widespread policy concern, there is little causal evidence on how shocks to teacher supply affect students' long-run outcomes. This paper provides the first such evidence by studying one of the largest teacher supply shocks in American history: the mass exodus of teachers during World War II. By the 1944–45 school year, roughly one-third of teachers had left the profession, and schools increasingly relied on emergency-licensed replacements.

To identify the effects of this shock, I exploit two sources of variation: geographic differences in the pre-war gender composition of teachers, and differences across birth cohorts in exposure to the wartime shortage. Because military mobilization drew overwhelmingly from male teachers, states with higher pre-war male teacher shares experienced larger supply declines. Using newly digitized data from the Biennial Survey of Education in the United States, I construct a measure of "missing teachers" that combines the national wartime decline with states' pre-war male and female teacher shares. The pre-war male teacher share is uncorrelated with mobilization rates, wartime industrial spending, and other state characteristics that could affect student outcomes. I then estimate a difference-in-differences model with two dimensions of variation. The first compares individuals born in states with more versus fewer missing teachers. The second compares cohorts who were school-aged during the war to older cohorts who had finished schooling before the war began. Outcomes and mechanisms are measured using pooled individual-level decennial census data from 1940 to 2020, complete-count census data from 1940 and 1950, and historical state-year records on staffing, salaries, credentials, and class size.

Greater childhood exposure to the wartime shock reduced both educational attainment and adult economic outcomes. A one-standard-deviation increase in missing teachers lowered high school completion by 1.6 percentage points (2.5%), any college attendance by 0.8 percentage points (3.5%), and college completion by 0.4 percentage points (3.6%). The same increase in missing teachers reduced adult weekly wages by 1.9%—approximately $38,000 in lost lifetime earnings for a full-time worker—and lowered the likelihood of holding a professional, managerial, or STEM occupation. Reassuringly, there are no effects on older "placebo cohorts" who had completed schooling before the war.

I find that more-affected states responded to the shock by hiring a larger share of emergency-licensed teachers, and these states experienced a persistent decrease in the share of teachers with a college degree. Class size also remained elevated in these states many years after the end of the war. The persistent decrease in teacher qualifications and increase in class size continued to negatively affect the educational outcomes and earnings of the next generation of children born after the war. These findings show that even short-term teacher labor market shocks can have long-run consequences for the teacher workforce and, in turn, for student outcomes.

Author

• Michael Briskin, Analysis Group
  Presenting Author