Opt-out or Dropout? Pandemic-era Public School Enrollment by Family Income

• In Event: Public School Student Disenrollment: Causes, Effects, and Potential Solutions

Abstract

Emerging evidence suggests public school enrollment losses during the COVID-19 pandemic persist. In this paper, we document the relationship between family income and public school disenrollment in Oregon. In particular, we seek to differentiate students whose families are opting out of public school for alternatives like homeschool or private school from those who are dropping out entirely, with attention to how these phenomena differ across the family income distribution.

We use tax returns filed in Oregon to define the school-aged child population on an annual basis, ensuring we do not conflate migration flows with changing enrollment decisions. Critically, these data include children not enrolled in public school, and provide a detailed measure of family income. Student records from the Oregon Department of Education provide information about which children enroll in public school (including charter and virtual options) in each school year.

We pursue several approaches to measuring enrollment in public school alternatives. First, in the public school data, we can analyze exit codes and high school graduation. Second, among all students, we can analyze linkages to 1098-T tax records which indicate tuition payments, providing insight into college enrollment rates. Third, among students linkable to the American Community Survey, we can analyze responses to the school enrollment question, though the sample size is limited.

We regress each outcome (public school enrollment, virtual school enrollment, college-going, etc.) on indicators for academic year interacted with a vector of family income bins, enabling us to trace out detailed enrollment patterns before, during, and after the pandemic. We have analyzed preliminary patterns and found substantial variation across grade levels. For example, kindergarten enrollments drop across the income distribution in 2021 and partially recover in 2022, while 12th grade enrollments exhibit substantial heterogeneity in income. Virtual schooling in 2021 may have delayed the substantial 12th grade enrollment drop for low-income students that appears in 2022, while middle-income families disenrolled immediately, and very high-income families held steady. We further document differences in enrollment responses based on the choice alternatives availble within communities.

Authors

• Leah Ruth Clark, U.S. Census Bureau
  Presenting Author

• Christopher Cleveland, Brown University
  Non-Presenting Co-Author

• Thurston Domina, University of North Carolina at Chapel Hill
  Non-Presenting Co-Author

• Paul Hanselman
  Non-Presenting Co-Author

• Andrew Penner, University of California, Irvine
  Non-Presenting Co-Author

• Winnie Yee, University of California Irvine
  Non-Presenting Co-Author

• Quentin Brummet, NORC at the University of Chicago
  Non-Presenting Co-Author

• Paul Youngmin Yoo, Stanford University
  Non-Presenting Co-Author

• Emily K Penner, University of California, Irvine
  Non-Presenting Co-Author