Impact of introducing a farm to school program on the number of school lunches served.

Farm to school (FTS) programs are becoming more prevalent throughout the United States. Yet, their impacts on students' behaviors are still not well understood. This study analyzes the impacts of introducing the local procurement aspect of a FTS program in a Florida school district on the number of National School Lunch Program (NSLP) meals served and the selection of salad meals prepared with FTS products using daily school-level point of sale (POS) data, product procurement records, NSLP menu data and linear panel regression analyses. The study district implemented the FTS program in 15 of its 22 elementary schools between mid-October of 2015 and mid-January of 2016. Weighted difference in differences regressions that account for the non-random assignment of the district's elementary schools into the FTS program and variation in the timing of implementation across schools indicate that the introduction of the FTS program did not affect the number of NSLP meals served or type of meals served at schools with FTS programs. We find some evidence that students selected more NSLP salad offerings on days when these meals were prepared with FTS products; however, these findings were no longer significant when school-level fixed effects were included in the model.