How food banks facing cuts could harness data, maximize efficiency
While federal food assistance cuts are too large to make up for in efficiencies, pooling resources among hunger relief agencies could help the remaining aid go further.
7 minutes
Experts
Julie Simmons Ivy
Chair of the University of Michigan’s Department of Industrial and Operations Engineering
Vivian L Carpenter Collegiate Professor of Industrial and Operations Engineering
Food banks often operate on the slimmest of margins as they help low-income families get the food needed. Recent proposed cuts to federal food programs such as the Supplemental Nutrition Assistance Program (SNAP) are putting an increased strain on the food distribution system and those who need assistance the most.
“People are going to see a reduction in their ability to have access to healthy foods when they are potentially food insecure, and I think that is the biggest concern for us,” said Elizabeth Hertel, director of the Michigan Department of Health and Human Services in a recent interview.
Julie Simmons Ivy, chair of the University of Michigan’s Department of Industrial and Operations Engineering, and Lauren Davis, a North Carolina A&T professor of Industrial and Systems Engineering, have collaborated for years on research designed to help food banks meet the demands of their communities. Their findings could help food banks make the most of their resources during a time of reduced supply and growing demand. Their methods could also help count the costs to communities from food aid cuts.
What are some of the operational challenges food banks face?
Ivy: Food banks are humanitarian supply chains connecting the food insecure with supplies from a variety of sources: manufacturers, retailers, growers and donations. It involves a network of partner agencies like food pantries and homeless shelters. The challenge isn’t simply supply or scarcity, but it’s logistics. Food often perishes before it can be used or it may get directed to places it’s not needed. However, as demand increases and supply decreases, preventing that will become more challenging.
Food banks try to provide equitable access to their supplies and try to do it efficiently. There are tradeoffs needed to ensure our food insecure neighbors, rural and urban, have access to the donated supply. In many cases, agency partners have been located with geographic concerns in mind—putting them in the center of a geographic area to try and make access equitable to the most people. In isolation, that approach is lacking.
And there is a degree of uncertainty with many parts of the system. Donations often aren’t constant. Transportation is a concern, not just with getting food from Point A to Point B, but getting people to the food or the agencies via public transportation. And volunteers are another key factor, because most food banks rely on them to be effective.
How does your research help deal with these kinds of issues?
Ivy: We take data from food bank partners related to their supply chain and enterprise resource planning systems. We look at what they distribute and how, and model the ways to increase efficiency and effectiveness equitably.
The data allows us to analyze what demand will look like over time. It identifies the network’s choke points, or capacity constraints. And we identify and predict vulnerabilities in the network, such as which agencies are in a strong position and which may see interruptions in service. Knowing that, partner agencies can be called on in times of need to supplement others experiencing shortages.
We can also analyze trends on the supply side, which can be uncertain because it’s largely donation based. Predicting the pounds of food donated over time enables identification of periods when a food bank receives donations above, or below, normal. This can guide product sourcing decisions, enabling a better matching of supply with demand.
Geographically, we can see if there are gaps in service, where they are located, rural or urban, and reassign resources to address them. For example, if certain rural areas are not getting the food they need, mobile pantries or other large-scale distribution efforts can be an answer.
Many food banks track their offerings by pounds of food. But what makes up those pounds? In many cases, areas getting less perishable food are also getting food that’s less healthy. We are examining how to get all areas the right mix of food. With our data, we can see what foods are exchangeable between food banks that could provide better balance. We can also maximize the speed of that sharing in order to help preserve perishable items.
What about situations that go beyond normal inconsistencies?
Davis: North Carolina is prone to hurricanes. When they occur, the food bank’s mission is unchanged even though the environment is more constrained. Partner agencies can be inaccessible or closed for long stretches, decreasing distribution capacity.For instance, Manna FoodBank, in North Carolina,experienced severe damage to its facility as a result of Hurricane Helene.
In our work with North Carolina food banks, we saw collaboration and resource sharing occurring among food banks to assist the population normally served by Manna FoodBank. But we also saw the potential for more robust preparation and planning, which our data-driven models could help enable.
For example, they can help with identifying the geographic areas and partner agencies most likely to be disrupted by a disaster and estimate demand for both food and non-food items. Informed by storm characteristics, the models can provide insight on how to strategically position and deploy resources—material, financial, and human—to efficiently and effectively meet the need during the immediate aftermath of the storm.
We’ve seen several cuts to federal food assistance programs this year, including the Local Food Purchase Assistance Cooperative Agreement Program (LFPA) and the Local Food for Schools Cooperative Agreement Program. How do these programs impact food banks, and how could your work benefit them?
Davis: Donations alone are not enough for food banks to meet their mission. The need for food is typically much higher than the supply. Programs like LFPA allow food banks to supplement donated food with more fresh produce options purchased from local farmers. As Julie mentioned, fresh food is more nutritious than most shelf-stable food, and many food banks are trying to make more of it available to their neighbors.
Programs like LFPA not only improve the health of individuals in need, particularly those with food restrictions, but they fuel the local economy, providing a stable market source for local farmers. Our work can help shed light on the impacts of cuts to these programs. In particular, we can quantify unmet demand resulting from reduced food availability, the potential reduction of healthy options available for distribution, and the overall impact on food bank operating budgets as they try to fill gaps opened by these cuts.
Food banks will have to make hard choices on how to meet the growing demand with fewer resources, which may necessitate an even greater reliance on data analytics and data-driven models.
Why is Industrial and Operations Engineering such an effective tool for dealing with problems like this?
Ivy: Industrial and operations engineering is focused on optimizing complex systems and processes. Hunger relief is a complex system, including donors, volunteers, federal and state food sources, partner agencies and food bank operations, all within a highly constrained environment. We map systems and the interactions between the elements that make them up—people, materials and information.
We are experts at using data analytics, decision modeling, optimization methods and quality improvement to understand the needs of a system. In the case of hunger relief, these tools show us how best to satisfy those needs equitably, efficiently and effectively. In particular, we center the human in all of our modeling and analysis.
And we know how to bring together the right expertise around complex problems.For instance, our team includes faculty and students from industrial engineering and operations research, computer science, civil engineering, and business.