Does School Starting Age Matter? The Impact of School on Childhood Obesity, Diet and Time Use in Australia

Tuesday, June 12, 2018: 1:30 PM
1055 - First Floor (Rollins School of Public Health)

Presenter: Ho-Ching (Angela) Mak

Discussant: Jonathan Cantor


Consistent with the global obesity epidemic, Australia’s rate of childhood obesity has shown an alarming upward trend. While a trend of this magnitude can only be explained by the environment, the exact mechanism remains unclear. Children spend about the same amount of time at home and in school, and therefore it may seem that the school environment is not responsible for childhood obesity. However, this paper reveals that the school environment is responsible for the phenomenon, and that the environment contributes mostly by exposing children to sugar-sweetened beverages, rather than by causing a lack of physical exercise. To establish this claim, this paper compares children who enter school early to children who enter school late, and finds large differences. In the presence of selection, early and late entrants may differ in their individual characteristics, and therefore a direct comparison of the two groups would not be appropriate. To address this identification issue, I implement a fuzzy regression discontinuity (RD) design, in which I exploit the continuity in these individual characteristics around the age cutoff for school entry. This design guarantees a fair comparison between the home and school environments’ contributions to childhood obesity. I apply this design to data from the Longitudinal Study of Australian Children (LSAC). My analysis reveals significant differences across the two groups. Early entrants are at least 26% more likely to be obese, and 22% more likely to have a waist-to-height ratio exceeding 0.5 (an indicator for
central adiposity). Early entrants are 19% more likely to be exposed to sugar-sweetened beverages. The two groups show no significant difference in exercise time, since exercise time with parents is largely substituted with that in school. This Australian case is of particular interest because the LSAC contains rich information on diet and time use, enabling this design to identify the exact channel behind childhood obesity in the school environment. This analysis complements the previous studies on this social issue.