Individual Submission Summary
Share...

Direct link:

School Calendars and Student Obesity: Evidence from California

Saturday, November 7, 10:15 to 11:45am, Property: Boston Marriott Copley Place, Floor: 3rd Floor, Room: MIT

Abstract

Childhood obesity is a major public health crisis in the United States, with prevalence quadrupling from 5% to 20% between the mid-1980s and 2020. Research consistently shows that obesity prevalence rises during summer vacation and falls during the school year, supporting the "structured day hypothesis" — the idea that schools promote healthier behaviors by regulating sleep, meals, physical activity, and screen time. Despite significant investment in school-based anti-obesity interventions, most programs have shown limited success. This raises an important and underexplored question: could changes to the school calendar itself affect student obesity? This paper estimates the causal effect of year-round school calendars on student BMI, overweight, and obesity. Year-round calendars do not add school days; they redistribute the standard 180 instructional days across the year, shortening summer break while adding shorter breaks in fall, winter, and spring. We examine whether these redistributions — and in particular, more disruptive "multi-track" year-round formats where students attend on rotating schedules — affect children's weight outcomes.

We use nearly two decades of data on California public elementary schools from 2000–01 through 2018–19. Student-level height and weight data come from restricted FitnessGram assessments administered to all California fifth graders, which we use to calculate BMI, BMI Z-scores, and overweight and obesity prevalence using CDC reference standards. Our analytic sample includes repeated cross-sections of fifth graders across a panel of 675 schools that switched calendar types during the observation period. We employ a staggered difference-in-differences design, comparing obesity trends in schools before and after calendar switches, using not-yet-switched schools as the control group. To avoid the well-documented bias of traditional two-way fixed-effects estimators, we implement two modern estimators: Chaisemartin and D'Haultfoeuille (2024) and Borusyak, Jaravel, and Spiess (2023). Models control for student race, ethnicity, gender, age, and school fixed effects, and distinguish between three calendar types: single-track, multi-track, and Concept 6 (a multi-track variant with fewer total school days).

Switching from a single-track year-round calendar to a traditional calendar produced no significant effect on any weight outcome. By contrast, switching from a multi-track calendar to a traditional calendar reduced average BMI by approximately 0.23 points (equivalent to roughly one pound) and reduced overweight and obesity prevalence by 1–2 percentage points, effects comparable in magnitude to those found in meta-analyses of school-based obesity interventions. Results for Concept 6 calendars were directionally consistent but inconclusive due to small sample sizes. The results suggest that what matters is not the length of summer vacation, but the extent to which calendars preserve or disrupt healthy routines. These findings offer a novel policy insight: calendar design choices made primarily for administrative reasons can have meaningful, unintended consequences for student health.

Author