The Economics of Retail Food Waste
96 Pages Posted: 28 Jun 2017 Last revised: 3 Jul 2019
Date Written: May 25, 2018
Each year, the perishable grocery industry produces substantial waste, resulting in considerable costs to firms and generating harmful greenhouse gas emissions. I study the incentives to produce such waste using a novel data set that matches a large supermarket chain’s loyalty-card database, including the time stamp for customer trips, with the firm’s product-level information on marginal cost, shelf life (perishability), and daily production-processes. In a descriptive analysis, I first show the firm’s production of waste correlates positively with the degree of demand uncertainty and with its market power. This relationship is consistent with the classic newsvendor problem and suggests positive waste levels are an endogenously-determined outcome. I then conduct a structural analysis of the artisanal bread category to assess the supermarket chain’s incentives to mitigate waste. Currently, grocery retailers have not yet adopted revenue management (intraday markdowns) due to insufficient scale. A hypothetical policy that reduces the costs of revenue management sufficiently to induce adoption would reduce total planned waste for the chain by 13%. Surprisingly, the incentive to reduce waste varies across the individual stores: in several stores, the firm endogenously increases planned waste, suggesting revenue management alone may be insufficient to curb grocery waste. I then simulate the effects of a ban recently enacted in California. The ban increases firms’ waste-disposal-costs by mandating composting. For the supermarket chain, this policy increases the returns to revenue management adoption from 7% to 11% and induces the firm to use intraday markdowns, resulting in a combined waste reduction of 31%.
Keywords: dynamic pricing, demand uncertainty, revenue management, perishability, retailing
JEL Classification: D61, D62, H21, H23, Q58, D81
Suggested Citation: Suggested Citation