Social and environmental analysis of food waste abatement via the peer-to-peer sharing economy

Global food waste is a significant problem, contributing to resource depletion and greenhouse gas emissions. In high-income countries, a substantial portion of this waste occurs at the post-retail and consumer levels. Food waste recovery is crucial for sustainability, and redistributing edible surplus food to those in need is a potential solution. The sharing economy, facilitated by peer-to-peer (P2P) platforms, offers a mechanism for efficient food redistribution. This study investigates the effectiveness and environmental implications of using a P2P food-sharing app (OLIO) to reduce food waste and address food insecurity. The researchers question whether sufficient interest from both providers and collectors exists, the overall environmental impact (including transport costs), and whether the platform benefits lower-income individuals experiencing food insecurity, given previous research showing higher-income groups often benefit more from sharing economy participation.

The introduction extensively reviews existing literature on global food waste, its environmental impact (8% of global GHG emissions, 20% of freshwater consumption, 30% of agricultural land use), and its economic cost (10–25% of household food expenditures). It highlights the disproportionate occurrence of food waste in high-income countries at the post-retail and consumer levels. The literature also points to the coexistence of food surplus and food insecurity in the same geographic areas, making food redistribution a viable strategy. The rise of the sharing economy and its potential for efficient asset sharing is discussed, alongside the challenges and uncertainties surrounding P2P food sharing, including the low economic value of food waste, potential supply and demand constraints, psychological barriers, and the potential for increased overall food consumption or negative environmental impacts from transportation. Existing research indicating a potential bias towards higher-income users in free sharing platforms is also noted.

The study uses a mixed-methods approach, analyzing data from OLIO, a popular P2P food-sharing app, between April 2017 and October 2018. The dataset included information on over 170,000 food item listings, including types, weights, retail values, collection success rates, and user characteristics. A supervised deep learning long short-term memory (LSTM) network was used to classify food listings into 13 categories. Monte Carlo simulations estimated the total weight and value of food diverted from waste, considering variations in weights across categories and user types. An environmental analysis focused on Greater London, comparing the life-cycle GHG benefits of avoided food waste with emissions from added transport. Six transport scenarios were modeled, considering different transport modes (car, London bus) and travel assumptions (one-way/two-way trips, walking for short distances). Socio-demographic analysis examined the income and education levels of providers and collectors using geographic census data. The study made several key assumptions; namely, that all multi-item listings were collected in full, notification location served as a proxy for collector's origin for travel distance calculation, general UK GHG emissions factor was applied for avoided food waste, and that food obtained via OLIO completely displaced new food purchases.

Over 19 months, 60% (90 ± 1t) of the over 170,000 food items listed on OLIO were successfully collected, representing an estimated retail value of £720,000-£750,000. Collection rates varied across food categories, with the highest rates for mixed listings (71%), sandwiches (70%), prepared food (66%), and fresh produce (65%). 'Food waste heroes' (OLIO volunteers collecting from businesses) were responsible for 71% of listings, with higher collection rates (66%) compared to regular users (47%). In Greater London, 41t of food waste were exchanged, resulting in avoided GHG emissions ranging from 87t CO₂eq to 156t CO₂eq, depending on the transport scenario. Even the most carbon-intensive scenario showed a net benefit, except for the most carbon-intensive travel scenario. Socio-demographic analysis revealed that most exchanges occurred between users in areas with lower income but higher education levels. Sensitivity analysis considered alternative emissions coefficients and the displacement rate needed for net environmental benefits. Even with lower substitution rates, net benefits were found to still exist, with benefits further increased with carbon opportunity costs of indirect land use changes.

The findings demonstrate the potential of P2P food-sharing platforms to divert significant amounts of food waste from disposal, particularly for short shelf-life items. The high collection rates indicate substantial demand for this food. The environmental analysis shows a net GHG benefit, even when accounting for added transport, emphasizing the platform’s positive environmental impact, especially with scenarios reflecting realistic London transport usage. However, there remains a need for scaling up P2P exchanges and addressing the potential for rebound effects through better understanding the displacement of new food purchases. The socio-economic analysis highlights a potential disconnect between the platform's users and those most in need of food assistance, suggesting food sharing might not fully address food insecurity issues. Additional limitations are noted.

This study demonstrates the potential of the P2P sharing economy to reduce food waste and generate significant environmental benefits. The high collection rates and net positive GHG balance highlight the platform's effectiveness. However, the study's findings also suggest that the platform may not be as effective at addressing food insecurity as initially hoped. Further research is needed to understand the limitations of the approach, explore strategies for expanding participation, and further investigate the potential rebound effects from increased consumer spending.

The study acknowledges several limitations. The lack of sub-listing level data introduced uncertainty in collection rates for multi-item listings. Using notification location as a proxy for collector's origin added uncertainty to travel distance estimates. A general GHG emissions factor was used, potentially underestimating the benefits for carbon-intensive foods. The assumption that food obtained through OLIO entirely displaces purchases might overestimate environmental benefits. Finally, the use of geographic census data to infer user income and education may lead to ecological fallacy.