By LeiLani Cauthen
Editor’s Note: This three-part series aims to identify the root problem to act on, rather than treating each challenge as isolated.
As noted in the first article, if the issue being addressed is not the real problem, it will not be solved—and it will likely worsen.
On the subject of student attrition, consider that up to 13.5 million K–12–aged children in the United States may not be participating in any formal schooling environment, including homeschooling. These students—who might be recent immigrants, part of transient families, or participating in emerging flexible learning arrangements—could be described as “off-grid.” American education data contains many gaps and inconsistencies. Large shifts are occurring, and most measurements are isolated datapoints or omit whole categories.
For example, a Gallup survey cited in news reporting suggested that the share of students enrolled in traditional public schools might drop substantially, which—given pre-pandemic public school populations—could imply millions of students leaving the system. Post-pandemic, districts nationwide continue to see higher-than-expected attrition. Enrollment losses translate directly into revenue declines, creating immediate budget pressures even as some federal stimulus funds temporarily offset losses for certain districts.
The Datapoint Issue
Here are representative datapoints from across multiple states and districts to illustrate the scale and breadth of the trend. No state or district appears immune. These examples show losses in both large and small systems; while a few districts or schools may experience growth, the broader pattern is of significant decline. Isolating single datapoints can mask a larger, systemic shift and lead decision-makers to underappreciate or misdiagnose the problem.
- New York State. City Department of Education enrollment was reported at just over 919,000 for 2021–22—a loss of nearly 37,000 students since the school year began and more than an 8% decline since the pandemic began. Officials cited the pandemic, population shifts, and declining birthrates.
- New York City. Approximately 75% of NYC public schools enrolled fewer students in a recent year, and nearly 23% of schools lost 10% or more of their students.
- Philadelphia. City budget projections have anticipated the loss of thousands of students—an expected 7,000 fewer students in a fiscal year would reduce staffing allocations and funding tied to enrollment.
- Arlington Public Schools, VA. About 2,000 students who left when buildings closed in March 2020 indicated they would not return.
- Del Mar Union School District, CA. K–6 enrollment dipped from 3,944 to 3,701 year over year.
- California (statewide). For the first time this century, fewer than six million students attend public schools. Statewide enrollment dropped by more than 110,000 students in one year, with charter enrollments also down and private school enrollment up modestly.
- Sarasota County, FL. Projected enrollment and actual counts diverged, costing the district millions in funding tied to per-student allocations.
- Chicago Public Schools. Analysis indicates CPS has lost tens of thousands of students over the past decade, leaving many traditional schools well under capacity while academic proficiency remains low in many grades.
These datapoints show substantial losses occurring at a scale the country has not experienced before. But datapoints alone are not the core problem—they are symptoms that point to deeper issues requiring investigation.
The Hidden Data Issue
Conflicting research and inconsistent reporting obscure the true magnitude and nature of changes in student participation. Several anomalies deserve attention:
- There are large disparities in estimates of the total population aged K–12 and the number actually accounted for in official counts—differences in the millions suggest significant undercounting or data classification issues.
- Surveys and administrative data disagree on homeschooling prevalence. For example, some federal sources reported that 5.4% of children were homeschooled in 2020–21, yet adult respondents reported higher household-level homeschooling rates. This discrepancy could represent millions of children not clearly captured in child-level datasets.
- New or hybrid learning arrangements make categorization difficult. Many offerings exist outside traditional public and private schools: online charters, private curriculum providers selling directly to families, subscription-based homeschool programs, and digital courseware with AI or live tutors. These alternatives often fall into gray areas that typical enrollment systems do not capture.
Put simply, the attrition many districts experience is real, but official reporting may understate, misclassify, or fail to capture where students are going. A growing number of families spend directly on digital curriculum, apps, and subscription services—consumer spending on these products increased as families invested in at-home learning resources. These consumer-directed options, combined with private and hybrid providers, create an education ecosystem only partially visible to traditional data sources.
The Vague Longing Issue
Beyond data gaps, there is a deeper cultural and practical shift. The pandemic accelerated an existing longing for change in American education. School leaders frequently report a sense that “something has changed” without being able to fully define it. Surveys of families leaving traditional schools indicate a large share choose “other” or “none of the listed reasons” when asked why—suggesting motivations that standard categories do not capture, including a desire for more flexible schedules.
Flexibility—especially control over time and place for learning—emerges as a central driver. Time is finite, and families value schedules that fit modern life. Space and the ability to be remote are also influential. Technology in other industries has transformed logistical experiences (ride-hailing, e-commerce, delivery, streaming), yet schooling largely retained a century-old structure: fixed schedules, physical buildings divided into classrooms, and a standardized conveyor-belt progression for all students.
The pandemic forced rapid experimentation with remote and hybrid models, and for many families that experience revealed alternative possibilities. Attrition is not merely movement toward existing alternatives; it is movement toward forms of learning that remain only partially defined—combinations of in-person and online, personalized content and pacing, and networks of human and digital supports assembled on demand. These hybrid models blend high-quality digital interfaces, improved curriculum workflows, and human oversight in ways that early distance-learning efforts often lacked.
Therefore, “back to normal” is a flawed aspiration. The new reality will likely be a spectrum between traditional schooling and fully online options—one that better integrates human guidance with robust digital systems, personalized learning experiences, and flexible use of time and space. Because the needs and preferences of learners are diverse, solutions will be plural rather than singular.
In short, attrition is a symptom pointing to structural, technological, and cultural shifts in how families approach learning. Addressing it requires rethinking how education systems measure participation, how they integrate digital and human resources, and how they accommodate flexibility and personalization.
About the author
LeiLani Cauthen is CEO and Publisher of The Learning Counsel. She has extensive experience in the digital content ecosystem, software development, adoption processes, and evolving models for school coverage. An author and media personality, she brings two decades of research, publishing, and market leadership in high technology, education, and government.
This article was originally published by The Learning Counsel, a research institute and news media hub focused on providing context for the shift in education to digital curriculum.