Gleaning research

Modeling the design space of gleaning in order to better inform design decisions for food security and other civic services

Gleaning research

Summary

Gleaning is the practice of salvaging food left over from its intended use. In this article, my advisor Dr. Carl DiSalvo and I present research into the activities of gleaning with an emphasis on the tools used in gleaning. From this research we identify a series of design opportunities. Perhaps the most fertile opportunities are related to socio-technical networking: the processes and infrastructures for providing information about the availability of food for gleaning and access to the actors who can move and store gleaned food.

Introduction

In 2013, an estimated 14.3% of Americans had inadequate access to food sufficient for an active and healthy life (Coleman-Jensen et al., 2014). Yet 31% of the food supply at the retail and consumer levels went uneaten (Buzby et al., 2014). This imbalance suggests inefficiencies in how the country’s food systems are managed on an institutional level. On an organizational level, food banks and similar entities are effective at reducing food insecurity by diverting some of the safe, edible food that would be wasted by grocers and restaurants to the organizations’ food-insecure clientele. However, these services are limited by their operating capacity. We found that on an individual level, people who pick up already wasted food from dumpsters and bins are limited by their means of transportation, knowledge of what food is still edible having been thrown away, and the social stigma of dumpster diving. These efforts by both organizations and individuals are known generally as “gleaning,” “food recovery,” “food rescue,” or “food salvage.” This variation in terminology is evidence of a spectrum of perspectives and activities. In this paper, “gleaning” refers to practices of salvaging food left over from its intended use.

Gleaning has long been a common post-harvest activity, where seasonal farm cycles would signal locals to glean leftover crops from the fields. Today over 80% of Americans live in urban areas (U.S. Census Bureau, 2012), away from not just the fields that produce food, but from the grocery stores that distribute it. Over 15 million low-income people in these urban areas live in “food deserts” where access to large grocery stores is limited, and many do not have access to a vehicle (Ver Ploeg et al., 2012). As our food systems take shape in more urban contexts, it becomes more appropriate to consider gleaning as a post-distribution activity.

Juxtaposition of gleaning then and now. Photos credited to William Vandivert and mrmondialisation.org, respectively.

Gleaning often occurs at large, community-size scales, with organizations like those interviewed in this survey and government programs providing for food security. Gleaning also occurs on strictly personal levels (Hoisington et al., 2001), such as bakers taking home unsalable pastries, dumpster diving behind a grocery store, and individuals living a freegan lifestyle based on sharing resources (Pentina & Amos, 2011). So while an individual may relieve one’s own food insecurity via these practices, it may not scale and solve insecurity on a community level, as Allen suggests. However, when these people’s individual interests are focused on activities related to shared problems like food insecurity, their practices can be analyzed as citizen politics—a conceptual framework for improving the effectiveness of public stakeholders in doing public work. By analyzing activities in this way, design opportunities can be more easily defined by common modes of public problem solving, including their strengths and limitations (Wagner, 1996). McCullum et al. found in one case that citizen politics was an effective framework for diverse community members to both find common ground on food security topics and build an agenda only when changes would not conflict with more powerful stakeholders (2002; 2003). When powerful government-affiliated stakeholders antagonize gleaning and community food security efforts, however, there is a challenge to Allen’s argument of community food security supplementing, rather than replacing, traditional government programs and policies. For example, Orlando, Florida enforces an ordinance that blocks any feeding event “likely to attract” 25 people to a public space without a permit, of which the city only provides two per year (§ 18A.09-2.). Similarly, corporate policy often prevents retailers from giving out unsold food and uses locked trash compactors instead of open dumpsters for disposal, despite national legislation encouraging and protecting donation (104th Congress, 1996).

Among other literature reviewed, there was no research that modeled the complex interactions between actors, policy, and the food itself. Whereas Allen (1999) suggested stronger interaction between community and government actors, McCullum (2001) cast doubt when power differences among the participants were substantial. Without better understanding how these factors interact and conflict, we are left with an inadequate picture of the design space for gleaning, which would seem to be a promising practice for addressing some issues in community food systems. Our research, then, was motivated by a hope to articulate this design.

In this paper we present research into the practices of gleaning, with an emphasis on the tools used in gleaning and how those tools work to structure a socio-technical system of exchange. From this research, we identify and discuss a set of themes and design opportunities.

Related Work

Over the past decade there has been an increased interest in food and food systems within the human-computer interaction design community. The approaches to food and food systems are diverse, expressing the pluralism of the HCI community. For some, cooking and eating (and to a lesser extent growing) are seen as opportunities for novel interactive systems and interaction techniques (Comber, et al 2012). More common, however, is the exploration of food and food systems in relation to health and sustainability. Given the increase in obesity and food-related disease, it is not surprising that many researchers in both HCI and various health fields see digital interventions as a promising site for tracking and changing eating behaviors (Andrew et al. 2013, Chang 2014). In many cases, food and food systems research has been framed in relation to sustainability. This work includes both efforts to increase awareness of sustainability issues with regards to food systems (Hirsch, et al 2010, Choi & Blevis 2010), efforts to support more sustainable food consumption (Bohner 2009, Clear 2013), and efforts to support more sustainable practices of growing food [Heitlinger, et. al 2013, Hirsch, et. al, 2010, Lyle, et. al 2015).

Within this corpus on food and food systems, we situate our research in relation to work on food waste (Comber 2013, Ganglbauer, 2013a, 2013a, 2014) and food justice (Grimes, 2008, Dombrowski 2012, 2013). Gleaning is a practice of making use of food waste, or rather, of keeping food from becoming waste through inventive means of recovery and use. To date, most studies of food waste and practices and tactics for recovering food waste have emphasized the domestic sphere and individual or small group dining. (Comber 2013, Dombrowski 2012, Farr Wharton 2014 Ganglbauer, 2014) Our work adds to that by exploring practices, tactics, and opportunities for food waste recovery at two other scales: the scale of the individual gleaner and institutional/organizational gleaning.

In addition, we draw inspiration from work exploring issues of food justice, or the intersection of food and food systems and social justice. (Grimes, 2008, Dombrowski 2012, 2013) As became apparent through our research, so much of the activities of gleaning are rooted in beliefs about sustainable and fair distribution and redistribution of resources. For many, gleaning is a way to work against injustices in the food system and contribute to the food security of a community. As such, emerging research at the intersection of food justice and HCI continues to motivate our work.

Methodology

We drew upon Ericksen’s (2007) framework for studying and evaluating food systems affected by issues of food security and social welfare to structure our investigation and analysis of gleaning practices. This framework helped guide data collection by defining and categorizing key activities and actors involved in food systems. For example, foraging is not gleaning, because there is no extant human intentionality behind food that could be foraged, whereas gleaning involves taking food that previously had an intended purpose. The framework also informed analysis by understanding the activities and actors individually, as well as how they interact in key processes for the system’s outcome. In addition, power imbalances McCullum et al. (2003) reported in setting food security agenda were important when analyzing the relationships between decision makers and among the scope of different policies.

Defining Activities

Ericksen’s framework groups food security activities into four categories: production, processing, distributing, and consuming. This concurs with a Natural Resources Defense Council report, which includes an additional category following production: postharvest handling and storage losses. Based on our review of gleaning organizations, we use these same frameworks to inform our own categories of gleaning activities:

  • Field Gleaning
    Harvesting crops left in the field or on the farm after harvest, often due to market forces and pest damage (Gunders, 2012). Large organizations like the Society of St. Andrews and government initiatives often operate within this category (Hoisington et al., 2001).
  • Small-Scale Transportation
    Often non-professional drivers of common vehicles like cars, vans, and bikes moving gleaned food between other activities. Volunteers drivers are critical to food pantries and peer-to-peer services like Nourish Now (Derrickson, Spellman, Rice, & Mahoney, 1999; Eisinger, 2002; Tarasuk & Eakin, 2003).
  • Large-Scale Transportation
    Large networks like Feeding America food banks require a reliable fleet of trucks to transport millions of pounds of food between sources, food banks, and agencies like food pantries.
  • Storage
    Long-term storage of food, often in bulk, often co-located or closely related to other activities. This includes food banks that manage large-scale gleaned food networks and small food pantries that distribute food to their clientele (Vitiello et al., 2013).
  • Food Pantries
    Food assistance organizations with facilities for acquiring donated food, especially from food banks, and distributing directly to individuals and families. This includes church kitchens, community food security organizations, small nonprofits, and homeless shelters (Vitiello et al., 2013).
  • Preparation for Consumption
    An end point to gleaned food is back on another person’s table after distribution by food pantries, dumpster diving, restaurants like wasteED, and stores like Daily Table.
  • Dumpster Diving
    Obtaining food from dumpsters for consumption is a common practice among homeless people (Eikenberry & Smith, 2005), freegans (Pentina & Amos, 2011), and even food-secure people (Vinegar et al., 2014), This activity reframes places of waste as places of second-hand resources.
  • Relationship Management
    Among all literature surveyed, networking is by far the most prevalent activity, though not itself a gleaning activity per se. However, it is a critical aspect of each category that enables disparate stakeholders and networks of stakeholders to work together. As such, it is important to consider this activity on its own as it relates to HCI.

Fieldwork

With these categories in mind, we sought out experts and prominent individuals involved in the services and organizations with public points of contact. We also identified some people through our own contacts and used our personal connections in addition to their public profiles to invite them to a semi-structured interview. We interviewed multiple people at larger organizations like a local food bank and food pantry, as they specialized in different gleaning categories and provided a more holistic picture of the organization’s role in gleaning in the local food system.

Each category’s social network in the Atlanta Metro Area was unique, making recruitment somewhat difficult. For example, food preparers could get in trouble for some gleaning practices, dumpster divers were only found by word of mouth, and all farmers surveyed had very little waste or surplus food to be gleaned. Therefore, for the latter, we interviewed a field gleaner from farms rather than farmers themselves. In total, we interviewed eight gleaning experts.

Participants Gleaning categories
Troy Alda, food sourcing specialists at the Atlanta Community Food Bank (ACFB) Large-scale transportation,

Storage,

Relationship management

Matthew Swift, food sourcing specialists at the Atlanta Community Food Bank (ACFB) Large-scale transportation,

Storage,

Relationship management

Ryan Watson, vice president of community services at the Atlanta Community Food Bank (ACFB) Relationship management
Richard Aniston, leader of Atlanta Food Not Bombs (AFNB) Food pantries,

Preparation for consumption,

Relationship management

Emma Simmons, former food pantry manager at Buckhead Christian Ministry (BCM) Food pantries,

Relationship management

Kim Stone, interim volunteer manager at Buckhead Christian Ministry (BCM) Small-scale transportation
Al West, a dumpster diver in Atlanta Dumpster diving
Taylor Bale, program coordinator for south Georgia at Society of St. Andrew (SSA) Field gleaning,

Small-scale transportation,

Large-scale transportation,

Relationship management

The bulk of the data presented in this paper comes from a series of interviews and follow-up emails on particular questions found from subsequent interviews. Each participant was invited to a semi-structured interview to discuss the gleaning practices of their organization, their own, and others they may be familiar with. In particular, participants were asked what enables and limits their current or planned activities so that this study could expose design opportunities. They were asked to describe the ways they glean, breaking down the individual tasks performed by the participant and their organization as a whole. They were also asked to speculate how they, their organization or community, and Atlanta overall could make them more effective at their gleaning activities and the issues they sought to resolve through these activities.

Findings

In what follows we present a description of the tools used to support gleaning across multiple organizations. Each organization operates at a different scale, with different commitments and practices, which is reflected in the tools they use, and do not use, to facilitate gleaning. Understanding tools uses within and across these organizations provides insight valuable to design by identifying what works and what needs are currently unmet.

Atlanta Community Food Bank

Feeding America was the U.S.’s leading domestic hunger-relief organization that supports a network of food banks like the The Atlanta Community Food Bank (ACFB) (Mills et al., 2014). Each food bank coordinated sourcing and distribution for its locale, improving food security by supplying both processed and fresh food to their agencies

As food sourcing specialists, Troy Alda and Matthew Swift’s goal was to maximize the ACFB’s food income from donors like farmers, food packers, and grocery stores. These food donations would be transported to the food bank’s storage facilities and then to its 600 agencies (including food pantries, soup kitchens, and other verified distributors) in 29 Georgia counties. Alda and Swift’s tasks were primarily related to communication and networking, and they indicated their most valuable tools were email and phone, both mobile and landline.

They also used software developed for Feeding America food banks called Choice and Produce Matchmaker. Choice was a web application that applied a market-like economic system to the food bank network and its sources, using food as the only tradeable items and “shares” as a fiat currency. Food sources electronically sent their donations to Feeding America, whose customer service department reviewed, processed, and posted to Choice. Once available there, food sourcers like Alda bid with their shares to win the items for their food bank. Shares were regularly allotted to food banks based on their needs, but they could also get more by bidding a negative number of shares on items. This often happened when an item may be undesirable or difficult to move, so a food bank charges shares (i.e. offers a negative amount) to take the item.

Choice used to be the system for both packaged goods and fresh produce, but in late 2015 a similar web application called Produce Matchmaker was developed to handle produce exclusively. Unlike Choice, food sources could directly upload their donations to the system, which was then vetted by Feeding America’s produce team. Another difference was that Swift said that the share-bidding system on Choice did not work out for produce, so food banks pay with money to transport produce claimed on Produce Matchmaker.

The ACFB website was another important digital tool for attracting visitors and providing information on how people could donate food to the food bank. Aside from ACFB’s contacts that he could contact directly, he could also network with contacts like produce distributor Fresh Point’s board of directors to seek out more donors. Similarly, organizations like the Georgia Food Industry Association maintained a list of businesses that he could cold-call or ask a mutual contact to introduce him to discuss donations to ACFB. For example, several times a year he would attend large events where there would likely be surplus food, such as the National Association of Convenience Stores’ trade shows. In this case, he would give sheets of stickers to attendees who were displaying their produce, asking them to attach the stickers to their unsold produce so ACFB could haul it away after the event.

The sourcing specialists worked closely with transportation and other logistics specialists at ACFB. ACFB relied heavily on their own fleet of trucks to move large amounts of food, usually from distributors (Kraft, ConAgra, Nestle, Kroger, and local warehouses) and retailers (Publix, Walmart, Sam’s Club, Kroger, Target, and BJ’s). To source food, Alda kept track of donors and potential donors, where food was and would be, and where the fleet drivers were. Without the fleet, donors and agencies would be far less accessible since they often lacked the time and resources to move the food themselves. Although, agencies also handled ACFB’s relationships and acted as “enabled” drivers to recover, store, and deliver donations. When ACFB’s trucks were unable to move a load of food, it was sometimes cost-effective to hire a commercial trucker. Other times when ACFB trucks weren’t immediately available, such as outside of business hours, he helped donors with how to optimize food recovery and storage until the trucks were available. Major distributors also use DonorExpress, also developed by the Choice creator, to make their donatable food available to Feeding America with less reliance on an individual food sourcer.

Finally, ACFB’s warehouse stored all their produce that couldn’t be immediately routed to their agencies. Though most of their donations were stored in pallets, they packed assortments of food for distribution to food pantries in Chiquita banana boxes, whose lids folded to leave a small opening in the top. This allowed the agencies to more easily check its contents to sort and distribute the food to their clients. Other food banks, Swift said, use similar second-hand items like milk crates; all of which are shareable on Choice. Since 2010, ACFB has almost doubled their annual donations to 60 million pounds; however, the warehouse’s pallet and freezer spaces were meant to store only 40 million pounds. Agencies, too, didn’t always have the logistical power for what they ordered. For example, a church operating as a food pantry may have limited kitchen space or van availability to store and move food as it becomes available. Agencies can use another product developed by the Choice creator called AgencyExpress to better handle these issues.

Atlanta Chapter of Food Not Bombs

Food Not Bombs (FNB) was an international organization that recovers food that would be discarded by stores and restaurants and shares it with communities to bolster their food security and to protest war and poverty, according to their website. Their local, often municipal, affiliates operate independently of each other, only having a name and general goals in common. Its website provided general information about what the overall mission and methods were for gleaning, including instructions for starting and contacting FNB groups. Related to the social stigma of certain gleaning activities, the website dispelled misinformation and linked to stories about individuals with the organization and its groups.

In Atlanta’s group (AFNB), started by Richard Aniston, members recovered food from their personal contacts in the food industry and warehouses. Every Sunday, they met at their house in the Edgewood neighborhood of Atlanta to cook and prepare the foods to freely distribute in Woodruff Park downtown. Their clients were often homeless or disabled, though passersby also lined up for a plate. Their house was a place to distribute food and supplies throughout the week, including anything that wasn’t consumed at the park, and to freeze food for distribution later. The group moved their food, table, cooking equipment, and other items in a van and another member’s pickup truck. Without their own vehicles, Aniston said that their work would be extremely difficult to coordinate. While at the park, the other members helped distribute food while Aniston worked to engage people in the area. In this case, Aniston brought drums that attracted a number of people passing by to join in temporarily and others for hours. This operation in a public park was an example of how FNB groups achieved their general goals of promoting awareness of homelessness and food insecurity.

Garnering attention like this helped build AFNB’s personal network of contacts that was critical to enabling AFNB’s operations. They tried using digital means, like text messaging and email lists, to find and manage contacts, but it did not help them coordinate both availability and pickup of food at the same time. That coordination was critical to gleaning foods with a small window of availability, such as from restaurants. Aniston suggested that they could have used a digital exchange board, similar to Craigslist’s free listings. However, he described Craigslist itself as “very alienating” since users would all be strangers, whereas AFNB’s contacts were all at least acquaintances or friends of friends. Instead, he was much more interested in systems that allowed AFNB to connect with people who were already part of a community. However, he stated that ACFB refused their request for food to distribute, citing a difference in organizational structures. He was leery of the degree of reliance on strangers for AFNB’s food network; friends of friends and people within a few degrees of separation were preferable to complete strangers.

Buckhead Christian Ministries

Buckhead Christian Ministries (BCM) was a non-profit organization working to prevent hunger and homelessness, partly through their food pantry primarily for low-income individuals and families, who they referred to as clients. Like Alda and Aniston, former BCM’s food pantry manager Emma Simmons managed food sourcing and storage. The food pantry was largely stocked with purchased items and food drive donations, but they also occasionally supplemented that with gleaned food. Kim Stone was the interim manager of BCM’s over 100 volunteers, four of which gleaned day-old baked goods from Publix grocery stores weekly. Simmons and Stone recalled three unaffiliated people who had donated baked goods every week: one gleaned from a local Costco, one from Starbucks and CVS Pharmacy, and another from an unknown source. Usually BCM only provided these gleaned goods as snacks in their lobby, but sometimes they added baguettes to clients’ packages or cakes to clients on their birthday. BCM’s policy on perishable food donations was not particularly strict but still liable for perishables, so they generally only received bread and pastries and eschewed meat and cheese.

BCM’s managers like Simmons and Stone relied on phone and email to handle most of their interaction with food sources like Publix. Simmons, however, said she hated working over the phone when she managed the food pantry. It often required a lot of time getting to the “right person,” such as the store manager, to coordinate their gleaning efforts through transfers and phone menus. These agreements were made on a more personal level rather than as an official feature of the store. Consequently, in Simmons’s experience, some contacts had dropped their (or their predecessors) agreement. While the agreements were in effect, contacts would package their goods in boxes and bags for BCM volunteers, with whom they had arranged for a pickup time (often in the morning). The volunteers would then load them into their car, sometimes to its limit, and drop it off at BCM. Simmons recalled sometimes volunteers were unable to glean all that was available from a source, and other times they would not want to due to the smell of the goods, such as onion bagels from a deli. On this particular donation of bagels, of which the deli packaged hundreds at a time in large plastic garbage bags, Simmons figured its packaging would be off-putting to clients. Instead, they made sure to only reveal them to clients when laid out on a tray as snacks or in clients’ packages.

Simmons suggested that a way of educating people on how to package goods for gleaning would be very beneficial to gleaners. She suggested new food licensees could be given guidelines on what they’re legally allowed to do with excess food. Food trucks, for example, could then act as both food source and gleaner, keeping food in a safe environment for delivery to a secondary distribution actor like BCM. For corporation and big companies like Publix, Simmons suggested they have better education about reducing their liability for gleaning practices and increase their own benefit. For example, donations could be logged as in-kind donations, which would be important for budgeting and being able to report that as a non-profit. Finally, Simmons suggested having easier ways to connect with food sources, as she had “no good channels to communicate or ask for [donating food].” In addition to a simple repository of food sources, she saw value in ways to codify interactions with those places. For example, she wanted a reliable way to get directly to the “right person,” since interacting over the phone was inefficient and frustrating for her. Still, this could not overcome not having enough volunteer gleaners available or store managers reneging on their gleaning agreement. She, like other interviewees, was confident that supply of salvageable foods was not a problem if one knows where to look. Having access to people with attitudes of anti-wastefulness and wanting to provide food for people with limited food access was a boon to making use of this knowledge. This didn’t extend to an organization like ACFB, from whom Simmons decided not to take food, because it had so many rules about what and how much BCM could give out, as well as how to qualify people to receive food. The two organizations didn’t have compatible philosophies, which in BCM’s case Laura described as, “ask as few questions as possible and say yes as often as possible.” This was similar to the case of Atlanta Food Not Bombs’ interaction with ACFB.

Dumpster Diving

Dumpster diving, also known as “skipping”, can be a secretive practice due to trespassing and property laws making it difficult to access dumpsters legally.

Al West frequently went dumpster diving in Atlanta with several communities. As a volunteer with local social justice non-profit organizations, West received a small stipend per month for food, but they found it preferable to dumpster dive to sustain themselves as a supplement. They lived in an intentional community where people lived together on the basis of shared intents and values, including sharing resources like food.. They would go dumpster diving with a group of at least two people, never alone. Even more impactful than just two people was having a whole community doing dumpster diving so that they knew who needed food or had ideas of how to get food. West referred to their community house as a hub for organizing dumpster diving outings, where they announced usually by word of mouth when they had food to share. If they had someone in particular in mind, they would text or call them.

There are digital tools like forums, maps, and wikis where organizations and individuals share spots for dumpster diving. West was not aware of these, but said they would have been useful. As Aniston of Atlanta Food Not Bombs put it, though, dumpster diving spots are like secret fishing spots and are not something divers would want to share publicly. West suggested that a way to have a secret group online where people could talk about organizing for dumpster diving trips, such as on Facebook, would have been useful. They supposed it could be open, though, to raise awareness and combat the stigma associated with the activity. West preferred Facebook, but they had seen more action on social media sites Tumblr, Instagram, and Twitter. They’ve also found zines on the subject, such as one how-to guide to dumpster diving written by a teenager who was doing quite well by the activity. Using a social network like these, West said, could have improved consistency in scheduling within and between their communities.

When they did organize, they found that having a car was perhaps the most important “technology.” They would drive to places they knew or drive around residential neighborhoods, looking for dumpster to dive for more than just food. West said it was a matter of keeping their eyes open and paying attention to their surroundings to find opportunities.  West knew someone who dumpster dived on foot, but they couldn’t carry much and taking public transportation is harder. To West, having a car was key to taking what their communities needed. Taking only as much as they needed was one of the first rules of dumpster diving, according to the people who helped and gave West examples of how to dumpster dive. They also taught West the order in which to check different places’ dumpsters, especially large grocery stores. Aldi’s was often their first stop, because there was no scent around the dumpster and it was easy to access. They also knew other divers would go there because they would find things left outside the dumpster for others to take, presumably because the previous divers had taken only enough themselves. Whole Foods, by contrast, was harder to access, because there was a fence surrounding the area with the dumpsters. Large grocery stores often also are locked trash compactors that can’t be accessed by dumpster divers. West said people also would be suspicious, thinking they were “hooligans” sneaking behind buildings. Suspicion and trust were major obstacles to remediating the stigma of dumpster diving. The houses they lived in was a community of people who also joined, so they did not have to worry about judgment or stigma. Anyone can do dumpster diving though, according to West. “A flashlight, [pocket] knife, dress appropriately and have a car and have a buddy,” West said, “That’s pretty much all you need. And some gumption.”

Society of St. Andrews

Society of St. Andrews (SSA) was a Christian organization somewhat similar to BCM but focuses more on field gleaning. Whereas other organizations are focused on gleaning as a modern post-distribution activity, gleaning from a field or from harvested but unprocessed produce is the traditional post-harvest method that has persisted for millennia. It also directly improves food security as the outcome is access to fresh and healthy produce and is directed primarily to people who are food insecure.

SSA operates as a national organization with state-level area coordination. For example, Florida, a state that produces much of the country’s produce, had five area coordinators and a state director. Taylor Bale, formerly one of those area coordinators, was the program coordinator for south Georgia. In this role, she found people interested in gleaning, especially large fields of produce, and planned gleaning opportunities. She also coordinated SSA programs like Harvest of Hope that focused on educating participants about local and global hunger issues.

Similar to Alda with ACFB, Bale was primarily a networker. The most limiting factor was SSA and farmers knowing the other exists. Bale’s first step was contacting farmers to explain SSA and then ask permission to bring volunteers to glean any food they may have. An important point to mention was that they’d be protected by volunteers’ liability waivers. Other farmers would also hear about SSA by word of mouth and would call them when they had truck loads of produce that had been refused by buyers. Even a small surface blemish caused by farm pests would reduce its attractiveness on store shelves and would not pass the standardized quality requirements for retail. Produce would be automatically and manually sorted into boxes for marketable produce, and imperfect produce would often go into a dump truck called “coal bins.” This was reportedly especially the case with green beans and potatoes, which many stores often subject to strict length and shape requirements. For large commercial farms, this could affect tons of produce per harvest. For smaller CSA farms and urban gardens, there was little to no waste that couldn’t be repurposed on the farm, such as for animal feed.

She often had only a few days to get the word out to farmers and volunteers and then to glean the food. Having access to communities and social networks was most enabling to this rapid networking task, though she admitted that she did not use social media as frequently as she could. Bale also considered SSA’s website highly useful to inform farmers about gleaning and to sign up volunteers to email lists for each area of the state, as volunteers usually didn’t want to travel more than an hour to glean. Rather than directly contacting each farmer, the optimal situation Bale said would be to establish long-term relationships and work with them to set up a gleaning calendar based on their growing and harvesting seasons. Having access to farming and gleaning communities helped her get groups together, mainly by email. Before those groups arrive to glean, Bale and the farmer would agree on a gleaning schedule and go to the farm to handle obstacles like drainage and safety issues and to determine what they would pick and how. When the gleaners arrived, they would distribute tools, pick produce from the field, and then load into trucks or individual volunteers vehicles, depending on the amount of food gleaned.

Once they finished on the farm, they would drive the produce to food banks, soup kitchens, and other agencies with loading docks. The state food bank association would help coordinate this networking step. One such place was a “food hub” that Bale described as “like a food bank on steroids.” Unlike other large food banks, it had sections for sorting, freezing, canning, and boxing produce, storage for these packaged goods, a commercial kitchen to prepare and teach people how to cook fresh produce, and loading docks for distribution via large- and small-scale transportation. It also acquired food from as far as North Carolina, compared to just the Atlanta Metro Area for ACFB.  Most food banks didn’t have these facilities for acquiring fresh produce, and those that did could be overwhelmed and challenged to distribute it while still within the “window of sustainability,” according to Bale. This was especially important in cases such as where a farmer donated a truckload (over 40,000 pounds) of cabbage rather than dumping it in the woods, as was common practice in rural areas where the cost to properly dispose of it was too high.

Finally, Bale would record the weight of all food gleaned by SSA, broken down by amount sent to each agency or taken by individual gleaners. Combined with other area updates, this information would then be compiled into reports in Microsoft Word format and submitted to the state director and then to the national office. The national office stored all these details in their database, which would then be used to generate reports and annual receipts to farmers for tax purposes.

Dicussion

General Tool Comparison

Gleaning is an activity that involves many digital and nondigital tools and services, used in Atlanta and beyond, for gleaning-related activities. Whereas much literature has examined non-digital tools, this study sought out digital media used in gleaning. We produced a series of diagrams to describe both the food systems under study and abstracted diagrams for comparison to other systems. There are dozens of digital tools, a list of which are compiled in Appendix A, including many maps explicitly for gleaning and websites with information related to gleaning. These tools are used for and between categories of gleaning activities, modifying how food flows between actors. This is pictured in Figure 1.

Figure 1: Flow direction of food between activities involving gleaning, clockwise from top-left: small-scale transportation, large-scale transportation, preparation for consumption, redistribution, field gleaning, dumpster diving, and storage. Relationship management is not pictured, but it is present in each flow.

Figure 1: Flow direction of food between activities involving gleaning, clockwise from top-left: small-scale transportation, large-scale transportation, preparation for consumption, redistribution, field gleaning, dumpster diving, and storage. Relationship management is not pictured, but it is present in each flow.

Appendix B provides a matrix view of this diagram using existing services as examples of flow, which Appendix C provides more insight to how technology is used by each service. The categories of gleaning activities derived from past studies and this study’s interviews can also be applied to these general (non-localistic) tools, as well. The most high-volume services and tools bridge multiple categories, including Food Cowboy, CropMobster, and Feeding Forward.

General Themes

Many tools are used between different activities, particularly in socio-technical networking and mobilizing small-scale transportation, suggesting that something usable in multiple, or changing, contexts would be valuable. In addition to tools, four general issues were common among the interview participants: outreach, mobilization, trust, and education.

Outreach

All interviewed participants seem to know there’s more food available “out there” to glean, but don’t always know how to confirm that or better reach out to others who may know more. Knowing that food is available is difficult. West was involved in several communities and groups in Atlanta, and whenever food was leftover from a meeting, they would use their personal contact lists to figure out what to do with it. West would even join church groups they weren’t formally affiliated with and take leftovers to people they knew, personally and incidentally, needed food and called it “paying it forward.” Knowing people personally was invaluable to both West and Aniston, who maintained relationships with people in restaurants, since they save food from being wasted before it ever makes it to the dumpster. Many of these dumpsters are locked or on private property, meaning accessing the dumpster can be prosecuted for trespassing. Reaching out to people in the food industry is important for mediators, like food banks. Alda and Bale made a full-time job out of just maintaining and seeking new relationships with food sources. However, in the gleaning food system, the secondary distributors like food pantries may not always have access to these mediators. In Aniston’s case, he wasn’t able to qualify as an agency for ACFB to distribute the food it recovers because AFNB is not formally organized, according to him. BCM chose not to reach out to ACFB as an agency, because they did not want to adopt their strict rules. Rules aren’t always barriers to outreach with organizations, though. In West’s case, they knew that their community used to have a relationship with a community-supported agriculture (CSA) group that would give a lot more food to their house than other clients. Normally CSA uses a shared risk model that incorporates more communal rather than capitalistic rules for its distribution policy. Bale, too, worked with community-based agriculture and hailed a recent urban agriculture ordinance (Atlanta City Council, 2014), all to better attract gleaners in this personal context. West was also involved in other communities dedicated to distributing food collected and prepared for distribution to local neighborhoods. This ability to connect to entire communities, though, were strongly linked to how well they could mobilize many people at once.

Mobilization

Whereas the interviewees all had their known pick-up, drop-off, and storage locations, they did not always have the means to move gleaned food from their source to someone who could use it. Instead, they were reportedly only able to mobilize people for certain legs of the food’s journey. For example, ACFB has many food sources throughout Georgia, one storage facility, and many agencies for distribution, and Alda said they often had to sacrifice some gleaning opportunities for others based on their truck fleet’s capacity and routes that day. Fresh produce was often handled far outside the city of Atlanta at the food hub Bale described, but SSA had to coordinate the entire state from one office. West didn’t have access to a car at their house for years after they started dumpster diving until a car-owning roommate joined. They also had a lot to say about what could have been if they could mobilize and schedule more people at once; however, their conflicting schedules and business often thwarted their ability to scale their efforts. In addition to organizations’ resources, the food itself has an expiration date. Even if there is a possible route over some time, the food must be kept from spoiling during this journey. The organizations with more support were able to support more centralized space to store food, from deep freezers to entire warehouses. Gleaners can be significantly limited by this logistical problem, although they were usually able to keep some of the produce they gleaned for themselves. This can be a powerful motivation, although they were still unlikely to participate if they had to travel far.

Trust

More fundamental than their abilities and goals, trust can influence a lot about these organizations and how non-members regard and interact with them. SSA’s faith-based mission likely enabled them to establish a personal relationship to farmers, communities, and individuals who share the same values. Whereas in AFNB’s case, their philosophy and affiliations with anti-authoritarianism worked well with people involved in social movements, but that could hinder relationships with formal organizations and government, despite their interest in always attracting people to join their cause and activities. Organizational structure, too, can clash with that of others, such as ACFB’s refusal to distribute its food to AFNB. Aniston himself was leery of trusting people they don’t know to certain degrees in their gleaning network. However, their preference for personally connecting with communities would allow for access to more people within fewer degrees of separation once trust is established with key contacts. Since these groups aren’t legally bound to their reasons for interacting (or not) with each other, establishing trust between organizations could be greatly influential and perhaps easy to facilitate to connect gleaning organizations and gleaners. In BCM’s case, store managers were critical to their ability to glean from a large enterprise like Publix. The support of a manager could mean the difference of a car full of bread every week. Considering corporate and individual stores’ policies, though, this could incur too much risk for both the store manager role and the gleaning organization’s reliance on this source. Similarly for individual contacts, like AFNB’s contacts at restaurants, if this trust and education isn’t maintained by the organizations, it may not be transferred to subsequent people in these roles. Among the communities West was in contact with, there was a strong trust among them about packaging and sharing gleaned food.

Education

Similar to issues of outreach and trust, educating people about gleaning is perhaps the most important. Wasting food is more common than gleaning it, and the legal grey areas reduce the visibility of gleaning. In the case of dumpster diving and preparation for consumption, it’s more of a health case, whereas safety is a top priority for gleaners on a farm and managers of large stores. Despite that, those who are both aware and capable of gleaning food require little information more than knowing what to do with salvageable food waste. Bale reported that once farmers understood SSA’s capabilities, they were quite receptive to field gleaning and often shared information with other farmers. Public perception of gleaning is another form of trust that could perhaps yield the greatest benefit in scale. If everyone were open to providing their surplus food to potential gleaners, then BCM’s deli contact may be more likely to provide bagels in a branded box instead of a garbage bag, West may have an easier time separating gleanable food from spoiled or rotten food, and Bale could better mobilize gleaners across the state. However, if the public perceives gleaners as untrustworthy or “hooligans” as West put it, then the stigma of both gleaning and making surplus food gleanable would likely persist. Similarly, West and their communities were taught by others the social mores of dumpster diving, mostly targeted toward other divers but also to the site of the dumpster. The dumpsters at Aldis were often their first stop for reasons described, and the etiquette of not leaving a mess and letting surplus food from one diver’s haul be more accessible to the next diver is an example. However, businesses that are perhaps over-cautious about dumpster diving could be made more aware that compactors and fences make dumpster diving an illegal or inaccessible activity. Simmonsn suggested that when businesses go through their initial food licensing procedure, they also get some educational information about donating their food, especially how to package it for donation. She said that even though information is available, managers were unsure whether and how to get involved with BCM without some convincing. This difference in whether an organization relies on an individual versus a role in a food source is most important when access to gleanable food is held by the latter.

These issues were common among the other tools and organizations surveyed (Appendix A). Food Not Bombs is a loosely organized national collective of local groups, so some issues may be more or less important in other cities. The same applies to food banks, church organizations. This discussion was centered on findings from Atlanta but could be applicable to other cities. Further investigation would be useful in connecting not just organizations within a city, but between cities as well. Although the reach of these organizations may not span past the city limits often, organizations like food banks and networking applications like Food Cowboy could be an important facilitator for increasing collective reach.

Design Implications

The problem space framed by this survey is the disparity between food access and food production. In this section, we scope out the implications for both the sociocultural interactions involving gleaned food and the human-food-interaction that Comber, et al. (2012) called for to address issues in their design space. The related works we cited call for digital platforms for food opportunities (Bohner et al., 2009; Lyle et al., 2015), mediating technologies for food system interactions (Clear et al., 2013; Hirsch et al., 2010), and tools for improving self-sustainability (Chang et al., 2014; Heitlinger et al., 2013). Combined with our findings, we suggest that most challenges in gleaning is related to networking issues. These issues have organizational, public, and personal design implications.

Organizational Internetworking

Organizations for gleaning try to compensate for the systemic failures of food systems that fail to provide security, often due to inadequate geographical and economic access for vulnerable populations. However, these organizations are also inadequate at fully compensating when they do not work together. They tend to form their own networks with both inclusive and exclusive criteria, with larger organizations like food banks tending toward inclusivity. This leads to disparate, even disconnected networks of actors seeking to improve food security. Zobel et al. (2016) found that technologies have been highly effective in other industries’ supply chains for coordinating and sharing information between entities, and could be just as valuable for networks of gleaners. Exclusivity is a major consideration to designing an effective solution to connect networks, which together could better mobilize resources and reach out to more potential gleaners.

More abstractly, there are issues and inefficiencies in node discovery and path cost optimization. Food is the information being shared in a peer-to-peer (P2P) food network, where nodes are actors and organizations that face some cost of moving food between each other. CropMobster and Food Not Bombs are constrained by geography and tight personal networks, respectively. Mesh networks (Lua et al., 2005) could eschew these constraints while preserving many of their benefits. For example, the messaging application FireChat uses a topology that accounts for a dynamic set of available peers to move information. As peers join and leave FireChat’s network, each message is still able to hop through them until it reaches its intended recipient. Similarly, a reliable set of organizations like food banks and their contacts form a predictable, semi-permanent network, whereas actors like dumpster divers and occasional volunteers provide more ephemeral connections within the network.

Public Networking

Helping to some degree here are dozens of tools and services that can organize and educate the untrained public to participate in gleaning. These have largely been relatively simple digitizations of previously non-digital technologies, like maps and information repositories. Many have also leveraged internet and inter-community networking technologies to increase the accessibility and speed of gleaning. A problem with gleaning network technologies in Atlanta is not the quantity of food to glean, people to mobilize, or places to find food, but the knowledge of all of three at the same time. Instead, for example, a gleaner may know when a grocery store usually tosses out certain food but not how much or whether there would be any left by the time they got there, reducing the ability to plan and act at critical times.

The proliferation of peer-to-peer civic technology (Knight Foundation, 2013) suggests there could be systemic optimizations in a digitally mediated network and user-centered design opportunities to contextualize the system to a public. In order to entice people to join this network, the participants in our study (and cited studies) seemed to rely on strong philosophical convictions to participate in food systems this way, as well as getting a cut of the gleaned items. They may have direct access to the gleaned items via field gleaning and dumpster diving, a free store like AFNB, or programs like SSA and ACFB. The three high-volume non-localistic services described in the general tool comparison section are also strongly marketed versus simpler platforms like LeftoverSwap. However, one design consideration here is trade-off between centralization and distribution of the service. These services all have the benefit of popularity through their managing companies, enticing more people to join and thus share food; however, they also may fail to localize their service to the nuances in individual food systems, making it vulnerable to the same inclusive and exclusive drawbacks previously described. Instead, it may be worth considering a distributed knowledge network that can be localized and customized but still be accessible by similar applications. This perspective is more of an “infrastructure as a service” (IaaS) rather than Food Cowboy and Feeding Forward’s “software as a service” (SaaS) (Manvi & Krishna Shyam, 2014; Rodero-Merino et al., 2010). We suggest further research into how concepts like these could apply to human-centered computing and low-tech organizations like AFNB and SSA.

Personal Networking

Ericksen (2008) called for greater understanding of individual agency in food system interactions, citing the inconsistencies in ecological, political, and food security goals. In addition to timing, mobilizing people in the first place could scale more easily if trust and education issues were resolved. Dumpster diving and explicitly leaving food out to be gleaned tend to raise suspicion, borne out of legal and safety precautions. However, as the participants and literature have related, these suspicions are poorly founded and at times counter-productive to gleaning practices of community food security.

By raising awareness through more widely trusted channels, like social media and more direct forms of ICT, it may be possible for a hotter medium to educate and engage non-gleaners. For example, a service that suggests picking up food from a common gleaning source and moving to a common destination, both on or near a route they frequent, would both directly contribute to gleaning efforts as well as more comfortably introduce non-gleaners to the practice.

Conclusion

The U.S. experiences a significant disparity between food production and food access, affecting millions of people, especially the poor. This difference is mitigated by gleaning, practices of which in Atlanta are described in this survey. The findings of our survey exposed issues and opportunities for design interventions in the interactions between activities and among actors. The most common issues in Atlanta are related to networking, particularly in discovering new sources and destinations of food and the capacity and availability of actors to move food between them. Since gleaning sources (e.g. grocery stores) and destinations (e.g. food pantries) are rarely co-located, moving food is almost always necessary and often occurs in amounts that require a vehicle. This leads to another issue of being unable to plan for and do gleaning when all knowledge of and capabilities to do so are available, leading to food waste even when the people involved intend to salvage it.

Further research may involve more in-depth case studies of organizations and tools described here to understand how they may be individually affected by networking technologies. We see value in developing a more robust model of gleaning by similar surveys in different cities’ food systems. Literature reviewed included location-specific nuances that could not be studied in Atlanta. Understanding both local gleaning efforts and more general gleaning practices would be essential to developing digital media that seeks to improve food security and prevent food waste as it spans disparate food systems.

My design work after this point is under the Gleanhub project.