Estimating Food Waste at the Individual Household Level

Abstract

We estimate food waste at the individual household level indirectly using a stochastic production frontier approach. The estimated average percentage of waste is about 30%-32%. In addition to a baseline model, we also develop a second model that tackle the issue of missing information on physical activities using a data imputation technique. Based on the results, we are able to explore the relationship between food waste and important demographic variables, and we find that household food insecurity, SNAP participation, and larger household sizes are associated with less food waste, whereas healthy diet practices and higher income lead to more waste.