Doing Good While Doing Well: Food Donations and Reverse Logistics
“It's the most wonderful time of the year
With the kids jingle belling
And everyone telling you be of good cheer
It's the most wonderful time of the year”
But that’s not the case for everyone, and unfortunately, there are many people in great need, both physically and mentally. And when it comes to food, there is a whole supply chain trying to help feed people in need. In the US, there are around 370 food banks, 200 of which are operating under Feeding America’s network, and distribute the food through 60,000 food pantries.
Where do these food pantries get their food from? In a recent research paper, John Lowrey and his co-authors, discuss the fact that the food banks mentioned above “recovered close to 1.7 billion pounds of food from retailers, eclipsing federal commodity donations (740 million pounds) by at least a factor of two.”
Long-time readers of this newsletter know that I frequently revisit supply-chain themes: (i) the amount of waste created in ensuring that customers have the desired product at the desired time ( through an increasing variety of products), and (ii) the fact that for most products, the reverse logistics supply chain dealing with their end-of-life, is very inefficient and cumbersome.
The relationship between these two themes is actually quite interesting when considering food. In their paper, John Lowrey and his co-authors examine the economic impact of food-bank donations on the performance of stores that donate food —performance in terms of their overall retail demand, pricing, and even store profit. In particular, they show that:
“... food bank donations serve to raise in-store food quality through a preemptive remove-and-replace strategy for perishables, which ultimately allocates greater shelf space to items with peak freshness, thereby increasing retailers’ capacity for fresh produce.”
And this is exactly what this post focuses on: While the social value of these practices is clear in terms of the supermarket chain’s image and its ability to include food donations and social good on its ESG reports, the economic value of donating food is less clear.
Before delving deeper into the mechanism and the decisions these stores make, let’s first outline the type of product we’ll focus on. The focal product is one where customers can discern the quality and proximity to perishability either through expiration dates (such as dairy products), or by closely inspecting the appearance (such as bananas).
Not all stores donate food, and even stores within the same chain don’t behave similarly. So let’s compare a store that donates to a food bank and one that doesn’t, and understand the main tradeoffs associated with that decision.
Newsvendors in Everything
Calculating how much to stock from each product is a newsvendor problem. There are two associated risks: underage (running out of stock and losing revenue) or overage (having too much inventory at the product’s end-of-life). The decision of whether to donate food or not, impacts the overage cost. But the way it impacts is not trivial. In fact, it looks like a mini newsvendor problem in its own right (as you know, everything in life is a newsvendor problem).
Let’s look at the two scenarios:
A store that doesn’t donate food. The manager can decide toward the end of a product’s life at two different points in time. The first threshold is when products are almost a week before their expiration or start looking “ripe,” and some customers are reluctant to buy them. At that point, and since the store isn’t donating, the products may need to be discounted, thus exhausting all possibilities of them being sold.
If the products remain unsold, they continue to degrade until they reach the second quality threshold, which is closer to the expiration date, and the retailer must discard them. But there’s a cost associated with discarding products since they can’t just be thrown in the trash. Stores usually have a contract with a waste management company, which charges a “tipping fee,” which can be substantial (e.g., $39.75 per ton for Franklin County).
So the overage cost is the lost margin due to discounting, and the cost of discarding the products.
A store that donates food. Since food banks need time to distribute the donated products, and don’t want those that stores are ready to discard, they are willing to take them at the first threshold (a week before expiration). So the store manager really has one decision point, which is when to donate. Let’s assume that they always donate at the first threshold.
What’s the cost of overage now?
Food banks come and collect the food, so there’s no need to pay the cost of discarding it, but the firm loses sales during the last few days between the first and second threshold. Since these are lower quality foods, we can assume that the likelihood these are purchased is lower than usual, but it still exists. Finally, there is also a small (or maybe not that small) tax benefit for donations, since they are deductible.
For their paper, Lowrey and his team use potatoes as a case study to estimate the tax benefits:
“Specifically, taking the retail profit margin for round white potatoes as the difference between the average retail price ($0.85/lb, U.S. Bureau of Labor Statistics) and the FOB wholesale price in Idaho ($0.34/lb, U.S. Department of Agriculture, Market News Service) in May 2022, the cost basis for our example item, therefore, is $0.34/lb, so that cost plus 50% of the retail profit margin results in a taxable base of $0.595/lb. Applying an average corporate tax rate of 15% to this total results in expected tax savings of $0.08925/lb in tax savings.”
Their back-of-the-envelope calculations show that the overage cost of discarding unsold potatoes (including the services of waste management and the lost taxable deductions) is around 21% higher than when donating.
The paper goes deeper, but I’m simplifying. As long as there are food banks you can donate to, and your customers care about the quality of the products they buy (and thus there is a dent in demand once they cross the initial quality threshold), you will have a lower overage cost, which means that you should donate.
Even More Benefits
And there’s more: Stores that donate products have a higher quality product.
In their study, the authors show that donations:
(1) raise the likelihood that consumers visit a donating store,
(2) increase demand for the products in donating categories,
(3) result in higher prices, and
(4) raise store profits.
In fact, their findings show that donations allow retailers to charge higher markups than they would be able to otherwise. The team’s counterfactual simulations show that donating stores see almost a 3% increase in equilibrium prices, and enjoy approximately 50% greater volumes compared to non-donating stores.
The mechanism is clear: retailers that donate can increase their stock levels (given the lower overage costs), while increasing their overall price (given the higher quality average product). Customers who value quality are more likely to shop there, resulting in higher prices and profits.
But food banks also need to distribute the products, so the actual donation process isn’t the last step of this reverse logistics supply chain.
One of my favorite work on the job market this year is by Sean Sinclair from Cornell University. Their work is inspired by a problem that the Food Bank for the Southern Tier of New York (FBST) faced while operating their mobile food pantry program —a service that many food banks expanded during COVID. In these systems, there’s a fixed route with multiple stops, and the challenge is deciding, upon arrival, how much food to allocate to each stop, without knowing the demand levels of any of the following stops.
The study’s main concern is the tradeoff between efficiency and fairness.
Efficiency means that all available food is allocated, in which case the solution is simple: allocate products to customers as they come. If there are multiple stops, allocate as many units demanded in each stop until the products are exhausted. The problem with this strategy is that it’s not fair. By the time the last stop is reached, there may not be any products left.
A fair solution would be to forecast how many stops there are and allocate the products proportionally based on that forecast. The problem is that this isn’t very efficient since there may be too much food left at the end, which will have to be discarded.
The paper shows that in such mobile settings, no solution is both efficient and fair. In fact, they show that any gain in one strategy is inversely proportional to losses in the other. But they provide an algorithm to achieve the points on the efficiency-fair frontier.
Good News Before the Holiday
But let’s get back to the stores. Given the economic benefits, one may wonder why don’t all stores donate? The answer is thatsome claim there are liabilities with donating food in close proximity to their expiration date.
Well… Good news! Congress just passed the Food Donation Improvement Act, which was introduced by Senators Pat Toomey (R-PA) and Richard Blumenthal (D-CT), and had a bipartisan coalition of representatives. One of the main benefits is expanding liability protection for food donations.
So to summarize, this is exactly what we need: market solutions that are also environmentally and socially optimal, where the legislature is working to remove friction.
Can we eliminate food waste? No
Can we eliminate inequality? Not in the short run
Can we try to find solutions? Yes.
Do we face tradeoffs? Yes.
The only constant in life is the search for tradeoffs in the endless newsvendor problems that we face.
Thanks for reading Gad’s Newsletter! Subscribe for free to receive new posts and support my work.