Working Paper: NBER ID: w11411
Authors: John Cawley; Chad Meyerhoefer; David Newhouse
Abstract: To combat childhood overweight, which has risen dramatically in the past three decades, many medical and public health organizations have called for students to spend more time in physical education (PE) classes. This paper is the first to exploit state PE requirements as quasi-natural experiments in order to estimate the causal impact of PE on student activity and weight. We study nationwide data from the YRBSS for 1999, 2001, and 2003 merged with data on state minimum PE requirements from the 1994 and 2000 School Health Policies and Programs Study and the 2001 Shape of the Nation Report.
We find that certain state regulations are effective in raising the number of minutes during which students are active in PE. Our results also indicate that additional PE time raises the number of days per week that students report having exercised or engaged in strength-building activities, but lowers the number of days in which students report light physical activity. PE time has no detectable impact on youth BMI or the probability that a student is overweight. We conclude that while raising PE requirements may make students more active by some (but not all) measures, there is not yet the scientific base to declare raising PE requirements an anti-obesity initiative.
Keywords: Physical Education; Youth Physical Activity; Childhood Obesity; Public Health
JEL Codes: I1; I2
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| state laws mandating minimum PE requirements (I19) | increased PE time (C41) |
| increased PE time (C41) | increased physical activity during PE classes (I19) |
| increased PE time (C41) | increase in days engaging in vigorous exercise (I19) |
| increased PE time (C41) | increase in days engaging in strength-building activities (I15) |
| increased PE time (C41) | decrease in days engaging in light physical activity (J22) |
| increased PE time (C41) | no statistically significant impact on youth BMI (I14) |
| increased PE time (C41) | no statistically significant impact on probability of being overweight (C46) |