Does Banning Carbonated Beverages in Schools Decrease Student Consumption?

In an effort to combat childhood obesity, many schools have banned the sale of carbonated beverages on school grounds. I evaluate the effectiveness of these measures by investigating their impact on household carbonated beverage consumption. I match households in Nielsen Homescan Data – a large data set which tracks tens of thousands of U.S. households’ grocery shopping trips a the bar code level - to their school district’s carbonated beverage policies for the period 2001-2009. I apply a triple differences framework, which exploits variation across three dimensions: 1) regional variation, as to whether the ban was implemented at the school district level; 2) variation across time, as the timing of the policies varied across school districts; and 3) variation concerning the age of household members, as children of different school levels were either included or not included in the policy. I find that when high schools ban the sale of carbonated beverages to students, households with a high school student experiencing the ban increase their consumption of non-diet carbonated beverages by roughly the equivalent of 3.5 cans per month. Increased consumption is greater in the quarter following the policy implementation and appears to persist even a year after the introduction of the ban. I present evidence that the average high school student consumes roughly 4.5 cans of non-diet soda per month in school, when carbonated beverages are available. Thus, the results suggest that the drop in student school consumption is substantially offset by increased household consumption. I argue that this compensation is only observed for high school students (as opposed to elementary or middle school students) due to the fact that it is in high school that the carbonated beverage bans were really binding and their implementation actually changed the content of vending machines in schools, while in the lower school levels it was frequent that carbonated beverages were not readily available even when a ban was not officially in place. The results shed light on the persistence of eating habits within the household, even following policy interventions intended to change these and provide a better understanding of the challenges policy makers can face when implementing policies intended to combat the obesity epidemic.