November 2021
From Farm to Kitchen
The Environmental
Impacts of
U.S. Food Waste
U.S. Environmental Protection Agency
Office of Research and Development
EPA 600-R21 171
-------
Disclaimer
This document has been reviewed in accordance with U.S. Environmental Protection Agency policy and approved
for publication. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
Authors
Kirsten Jaglo, ICF Incorporated, LLC
Shannon Kenny, U.S. EPA Office of Research and Development
Jenny Stephenson, U.S. EPA Region 9
Reviewers
EPA would like to thank the following people for their independent peer review of the report:
Abhishek Chaudhary, Ph.D., Indian Institute of Technology (Kanpur, India)
Roni Neff, Ph.D., Johns Hopkins University (Baltimore, Maryland, USA)
Xiaobo Xue Romeiko, Ph.D., University at Albany, State University of New York (Rensselaer, New York, USA)
Acknowledgements
EPA would like to thank the following researchers for providing additional details related to their published
research: Catherine Birney, Xuezhen Guo, Marco Pagani, and Quentin Read. EPA would also like to thank Tim
Torma and Alexandra Stern for their contributions to the report.
This report was prepared with support from ICF Incorporated, LLC, under U.S. EPA Contract No.
68HERC19D0003. The external peer review of the report was coordinated by Eastern Research Group, Inc.
under U.S. EPA Contract No. EP-C-17-017.
Disclaimer and Acknowledgements
-------
Executive Summary
Over one-third of the food produced in the United States is never eaten, wasting the resources used to produce it
and creating a myriad of environmental impacts. Food waste is the single most common material landfilled and
incinerated in the United States, comprising 24 and 22 percent of landfilled and combusted municipal solid waste,
respectively. This wasted food presents opportunities to increase food security, foster productivity and economic
efficiency, promote resource and energy conservation, and address climate change.
As the United States strives to meet the Paris Agreement targets to limit the increase in global temperature to 1.5
degrees above pre-industrial levels, changes to the food system are essential. Even if fossil fuel emissions were
halted, current trends in the food system would prevent the achievement of this goal. Globally, food loss and
waste represent 8 percent of anthropogenic greenhouse gas emissions (4.4 gigatons CC>2e annually), offering an
opportunity for meaningful reductions.
Reducing food waste can also help feed the world's growing population more sustainably. The United Nations
(UN) predicts that the world population will reach 9.3 billion by 2050. This population increase will require a more
than 50 percent increase in food production from 2010 levels. Decreasing food waste can lessen the need for new
food production, shrinking projected deforestation, biodiversity loss, greenhouse gas emissions, water pollution,
and water scarcity.
In 2015, the United States announced a goal to halve U.S. food loss and waste by 2030, but the nation has not
yet made significant progress. The U.S. Environmental Protection Agency (EPA) prepared this report to inform
domestic policymakers, researchers, and the public about (1) the environmental footprint of food loss and waste
(FLW) in the U.S. and (2) the environmental benefits that can be achieved by reducing U.S. FLW. The report
examines the farm-to-kitchen (cradle-to-consumer) impacts of FLW, excluding the impacts of managing FLW
(e.g., methane emissions from landfills), which will be covered in a separate companion report (The
Environmental Impacts of U.S. Food Waste: Part 2).
Given the size and dynamic complexity of the U.S. food system, no single agreed-upon comprehensive estimate
of the total amount of U.S. FLW exists. Instead, the literature includes multiple credible estimates, which differ in
scope and methodology, that together provide insights into the magnitude and distribution of U.S. FLW. Estimates
that include food lost or wasted during all stages of the food supply chain (from primary production to
consumption) range from 73 to 152 million metric tons (161 to 335 billion pounds) per year, or 223 to 468 kg (492
to 1,032 pounds) per person per year, equal to approximately 35 percent of the U.S. food supply. Roughly half of
this food is wasted during the consumption stage (households and food service), and fruits and vegetables and
dairy and eggs are the most frequently wasted foods.
This uneaten food results in a "waste" of resources—including agricultural land, water, pesticides, fertilizers, and
energy—and the generation of environmental impacts—including greenhouse gas emissions and climate change,
consumption and degradation of freshwater resources, loss of biodiversity and ecosystem services, and
degradation of soil quality and air quality. Each year, U.S. FLW embodies:
• 560,000 km2 (140 million acres) agricultural land - an area the size of California and New York combined;
• 22 trillion L (5.9 trillion gallons) blue water - equal to annual water use of 50 million American homes;
• 350 million kg (778 million pounds) pesticides;
• 6,350 million kg (14 billion pounds) fertilizer - enough to grow all the plant-based foods Americans eat
each year;
• 2,400 million GJ (664 billion kWh) energy - enough to power more than 50 million U.S. homes for a year;
and
• 170 million MTCChe GHG emissions (excluding landfill emissions) - equal to the annual CO2 emissions of
42 coal-fired power plants.
This uneaten food also contains enough calories to feed more than 150 million people each year, far more than
the 35 million estimated food insecure Americans. To estimate the environmental impact of FLW, researchers
consider the amount of food lost or wasted as well as the type of food lost or wasted and supply chain stage at
Table of Contents
ii
-------
which it was lost or wasted. Food wasted further along the supply chain carries more impacts than food lost or
wasted earlier, since the impacts are cumulative. For example, food lost during primary production embodies the
resources used to grow the food, whereas food wasted during the consumption stage embodies the resources
used to grow, process, package, store, and distribute the food up to the point the food reaches the consumer.
Given the substantial environmental impacts of FLW, halving FLW- as the U.S. aims to do - could meaningfully
reduce the resource use and environmental impacts of the U.S. food system. Researchers estimate that halving
U.S. FLW could reduce the environmental footprint of the current cradle-to-consumer food supply chain by:
• More than 300,000 square km2 (75 million acres) agricultural land - an area greater than Arizona;
• 12 trillion L (3.2 trillion gallons) blue water - equal to the annual water use of 29 million American homes;
• Nearly 290,000 metric tons (640 million pounds) of bioavailable nitrogen from agricultural fertilizer with the
potential to reach a body of water, cause algal blooms and deteriorate water quality;
• 940 million GJ (262 trillion kWh) energy - enough to power 21.5 million U.S. homes for a year; and
• 92 million MTCC^e GHG - equal to the annual CO2 emissions from 23 coal-fired power plants.
Note that these estimates are conservative in comparison with other published studies presented in this report,
and that these savings can only be achieved through prevention (i.e., source reduction) of FLW. Recycling of food
waste cannot achieve these benefits since a substantial fraction of the impacts occur during the primary
production of food.
Modeling in the scientific literature also offer insights into how to maximize the environmental benefits of FLW
reduction programs and policies, which the report summarizes into three key points:
1. The greatest environmental benefits can be achieved through prevention rather than recycling.
2. The largest energy and greenhouse gas emissions benefits can be obtained by reducing FLW from
households and restaurants.
3. Focusing on reducing FLW of the most resource-intensive foods, such as animal products and fruits and
vegetables, can yield the greatest environmental benefits.
The report also examines U.S. FLW in global context to evaluate the U.S. contribution to this global issue and to
highlight key similarities and differences among regions and countries. Currently the United States wastes more
food and more food per person than most any other country in the world. Also, the environmental impact of each
unit of U.S. food loss and waste is greater than that of most other countries, as the U.S. wastes more food
downstream and more animal products than the global average. Fortunately, positive examples of progress are
emerging in similar countries. Over the last decade, countries such as the United Kingdom and Japan have
substantially reduced food waste, contributing to the global effort under the UN Sustainable Development Goals.
As global populations and incomes rise, and the environment faces pressures from increased food production,
reductions in the per person environmental footprint of agriculture will be essential to the sustainability of the
planet. Limited options are available to sustainably increase the global food supply to meet growing demand.
Closing yield gaps and increasing productivity alone will likely be insufficient to prevent further deforestation and
environmental degradation. Even under the most promising scenarios of yield increases, up to 20 percent more
land will be needed by 2050. Thus demand-side measures, such as reducing FLW or dietary shifts, will also be
needed to sustainably increase the food supply. A recent study projects halving global FLW could result in a 24
percent reduction in cumulative global food system greenhouse gas emissions between 2020 and 2100 (331 Gt
CChe), compared to a business-as-usual scenario. Significant reductions (6 to 16 percent) could also be achieved
in the amounts of agricultural land, water, and fertilizer used in 2050 (compared to business-as-usual scenario) by
halving global food loss and waste.
Key research needed to help the United States meet its goal to halve food loss and waste includes:
• Enhancing the data on U.S. FLW by improving precision and addressing data gaps.
• Increasing frequency at which the United States can track progress in reducing FLW.
• Quantifying the environmental impacts associated with U.S. waste of imported foods.
• Strengthening understanding of the interaction among food system supply chain stages with regard to FLW.
• Evaluating the life cycle impacts of proposed FLW prevention strategies.
• Exploring how trends in the U.S. food system will affect FLW and its environmental footprint in the future.
• Deepening our understanding of drivers of FLW unique to the United States.
Table of Contents
iii
-------
Table of Contents
Chapter 1. Introduction 1
1.1 Purpose 1
1.2 Background 2
1.3 Scope and Definition of Terms 3
1.4 Report Overview 3
1.5 Methods 4
Chapter 2. Environmental Footprint of the U.S. Food Supply Chain 5
2.1 U.S. Food Supply Chain 5
2.2 Environmental Footprint 6
2.3 Inputs and Environmental Impacts 7
2.4 Imports 11
2.5 Other Factors 13
Chapter 3. Characterization of U.S. Food Loss and Waste 14
3.1 U.S. Food Surplus 14
3.2 Total U.S. FLW 16
3.3 U.S. FLW Over Time 19
3.4 Edibility of U.S. FLW 20
3.5 U.S. FLW, by Supply Chain Stage 22
3.6 U.S. FLW, by Food Category 25
3.7 Measurement Methodologies 28
Chapter 4. Environmental Footprint of U.S. Food Loss and Waste 31
4.1 Methodologies 31
4.2 Agricultural Land Use 34
4.3 Water Use 37
4.4 Pesticide and Fertilizer Application 40
4.5 Energy Use 44
4.6 Greenhouse Gas Emissions 47
4.7 Summary of Environmental Footprint 52
Chapter 5. Environmental Benefits of Reducing U.S. Food Loss and Waste 56
5.1 Environmental Benefits, Relative to Current Footprint 56
5.2 Environmental Benefits, Relative to Future Footprint 59
5.3 Summary of Environmental Benefits 61
5.4 Environmental Benefits, By Food Category and Supply Chain Stage 62
Table of Contents ^¦ iv
-------
Chapter 6. U.S. Food Loss and Waste in Global Context 63
6.1 U.S. Share of Global FLW 63
6.2 Share of Food Supply Lost or Wasted 64
6.3 Characterization of FLW 65
6.4 Per Person FLW 67
6.5 Progress Towards Reducing Per Person FLW 70
6.6 Global Environmental Footprint of FLW 70
6.7 Environmental Benefits of Halving Global FLW 74
Chapter 7. Conclusions and Research Gaps 79
7.1 Conclusions 79
7.2 Strategies to Maximize the Environmental Benefits of Halving U.S. FLW 81
7.3 Research Gaps 82
References 84
Appendix A: Inputs and Environmental Impacts 94
Appendix B: USDA and FAO FLW Data 101
Appendix C: Literature Search Methodology 103
Table of Contents v
-------
CHAPTER 1.
Introduction
Over one-third of the food produced in the United States is never eaten, wasting the resources used to produce it
and creating a myriad of environmental impacts (FAO, 2019b; CEC, 2017). Food waste is the single most
common material landfilled and incinerated in the United States, comprising 24 and 22 percent of landfilled and
combusted municipal solid waste, respectively (U.S. EPA, 2020f). This wasted food presents opportunities to
increase food security, foster productivity and economic efficiency, promote resource and energy conservation,
and address climate change.
As the United States strives to meet the Paris Agreement targets to limit the increase in global temperature to 1.5
degrees above pre-industrial levels, changes to the food system are essential. Even if fossil fuel emissions were
halted, current trends in the food system would prevent the achievement of this goal (Clark et al., 2020). Globally,
food loss and waste represents 8 percent of anthropogenic greenhouse gas emissions (4.4 gigatons CC>2e
annually) (FAO, 2015b), offering an opportunity for meaningful reductions.
Reducing food waste can also help feed the world's growing population more sustainably. The United Nations
(UN) predicts that the world population will reach 9.3 billion by 2050. This population increase will require a more
than 50 percent increase in food production from 2010 levels (UN, 2020a; Searchinger et al., 2019). Decreasing
food waste can lessen the need for new food production, shrinking projected deforestation, biodiversity loss,
greenhouse gas emissions, water pollution, and water scarcity (Springmann et al., 2018; Jalava et al., 2016;
Bajzelj et al., 2014; Kummu et al., 2012).
In 2015, the United States announced a goal to halve U.S. food loss and waste by 2030, but the nation has not
yet made significant progress. Currently the United States wastes more food per person than most any other
country in the world1 (Chen et al., 2020). Over the last decade, countries such as the United Kingdom (UK) and
Japan have substantially reduced food waste, contributing to the global effort under the UN (WRAP, 2020; Parry
et al., 2015). Halving U.S. FLWcan help tackle climate change, feed those in need, and protect water availability,
water quality, air quality, and biodiversity and ecosystem services.
1.1 Purpose
The U.S. Environmental Protection Agency (EPA) prepared this report to inform domestic policymakers,
researchers, and the public about (1) the environmental footprint of food loss and waste (FLW) in the U.S. and (2)
the environmental benefits that can be achieved by reducing U.S. FLW.
This report provides estimates of the environmental footprint of current levels of FLW to assist stakeholders in (a)
clearly communicating the significance of FLW; (b) decision-making among competing environmental priorities,
including FLW; and (c) designing tailored FLW reduction strategies that maximize environmental benefits. The
report also identifies key knowledge gaps where new research could improve our understanding of U.S. FLW and
help shape successful strategies to reduce its environmental impact.
As shown in Figure 1-1, this report (The Environmental Impacts of U.S. Food Waste: Part 1) examines the
environmental burden of the "cradle-to-consumer" (i.e., "farm-to-kitchen") segments of the U.S. food supply chain
- beginning with primary agricultural production and continuing through distribution, processing, and retail, ending
with the consumption (or waste) of food at home or away from home, such as at restaurants and cafeterias. A
companion report (The Environmental Impacts of U.S. Food Waste: Part 2) will examine the environmental
footprint of the pathways for food once it becomes "waste" (i.e., is not consumed), such as landfilling, combustion,
composting, and anaerobic digestion. Together these two reports encompass the net environmental footprint of
U.S. FLW.
1 The U.S. wastes more food per person per day (measured in calories) than any other country and wastes the third largest
amount of food per person per day (measured in grams) behind New Zealand and Ireland (Chen et al., 2020).
Chapter 1. Introduction
1
-------
Cradle-to-Consumer Food Supply Chain
Primary
Production
Distribution
& Processing
Retail
Consumption
Food Waste
Management
FIGURE 1-1. STAGES OF THE U.S. FOOD SUPPLY CHAIN
1.2 Background
Recognizing the critical importance of reducing food waste, in September 2015, the United States announced the
U.S. 2030 Food Loss and Waste Reduction Goal to halve per person food waste at the retail and consumer level
by the year 2030 (U.S. EPA, 2020d). This goal aligns with Sustainable Development Goal (SDG) Target 12.3 of
the 2030 Agenda for Sustainable Development, a wide-ranging resolution adopted by the UN General Assembly
in October 2015 (UN, 2015). National governments representing roughly half the world's population have adopted
comparable food waste reduction goals (Flanagan et al., 2019). SDG Target 12.3 also includes the goal of
reducing food losses during production, though a quantitative target was not set.
Achieving this goal is ambitious. Fortunately, many U.S. states, municipalities, institutions, and private sector
businesses have established food waste reduction goals or initiated programs to reduce food waste in recent
years. Laws and executive orders in at least three states (New Jersey, Oregon, and Washington) have
established goals to reduce food waste by half by 2030 (NCSL, 2020; State of Oregon, 2020; State of New
Jersey, 2017). At least seven states (California, Connecticut, Maryland, Massachusetts, New York, Rhode Island,
and Vermont) have enacted organic waste recycling laws, most of which apply to waste by large commercial
sources (Maryland DOE, 2021; ReFED, 2021 b; Heller, 2019). On July 1, 2020, Vermont became the first state to
institute a statewide ban on sending residential food scraps to landfills. Cities such as Baltimore and Denver have
developed food waste reduction and recovery strategies, and a growing number of cities, including Austin,
Boulder, Minneapolis, New York City, San Francisco, and Seattle, have organic waste bans or organics recycling
programs in place (ReFED, 2021 b; Heller, 2019). Two-thirds of the world's 50 largest food companies have set a
FLW reduction target consistent with SDG Target 12.3 (Flanagan et al., 2019); and 34 businesses and
organizations have publicly committed to halve FLW in their U.S. operations by 2030 as part of EPA's Food Loss
and Waste 2030 Champions group (U.S. EPA, 202Qd).
Chapter 1. Introduction
2
-------
The annual market value of U.S. FLW is estimated to be $408 billion (ReFED, 2021 a). While actions to prevent
FLW typically carry some cost (i.e., time or money), reducing FLW can lead to financial savings for households
and businesses (Champions 12.3, 2017a; WRAP, 2013). For example, an investment in a food waste education
campaign of 26 million British pounds (GBP) over five years in the UK resulted in an estimated savings of 6.5
billion GBP—and 3.4 million metric tons of carbon dioxide equivalent (million MTCChe) in annual avoided
emissions—from households wasting less food (Champions 12.3, 2017a). A recent analysis by ReFED, a multi-
stakeholder nonprofit organization, projects the United States could achieve net financial benefits from a wide
variety of food waste prevention strategies, including enhancements to demand planning, packaging, surplus and
imperfect produce channels, inventory management, date labeling, donation infrastructure, and education
campaigns (ReFED, 2021a). A study sponsored by Champions 12.3 (a coalition of senior executives from
business, government, international organizations, and research institutions) examined food waste reduction
efforts in 700 companies (1,200 sites) across 17 countries including the United States. The study found that more
than 99 percent of the sites had a net positive financial return from their investment in food waste reductions and
that the median benefit-cost ratio was 14 to 1 (Champions 12.3, 2017a). In addition, by reducing FLW, the
substantial cost of disposing of FLW could be reduced. A recent Commission on Environmental Cooperation
(CEC) report indicates FLW accounts for $1.3 billion in tipping fees (charges for waste disposal) in the United
States (CEC, 2017).
While some FLW is necessary for food system resilience (i.e., intentional redundancy to prevent shortages or
price shocks with unexpected weather or natural disasters), and some may be unavoidable, broad consensus is
forming around the ability and need to halve FLW (ReFED, 2021a; Bajzelj et al., 2020; FAO, 2020).
1.3 Scope and Definition of Terms
This report focuses on the environmental impacts of producing food that is ultimately wasted. Food waste also
has important and far-reaching social and economic impacts. Although these impacts are outside the scope of
this report, they are discussed briefly when they intersect with environmental issues or to provide relevant context.
In this report, the term "food loss and waste" (FLW) is defined as food intended for human consumption that is not
ultimately consumed by humans. Information about food grown for other purposes, such as biofuels or feed for
animals not raised for human consumption, is excluded. However, when estimating the environmental footprint of
producing livestock and farmed seafood, data on animal feed is included when possible. Food donated to food
banks or upcycled into new food products is considered "surplus" or "excess" food and not FLW. Food that is
recycled or disposed is considered FLW.
The terms "food loss and waste," "food waste," and "wasted food" are used interchangeably to describe food loss
and waste in the report. Generally, studies on FLW define "food loss" as food that is not consumed due to
unintentional limitations in production or supply. For example, food might be left unharvested or unutilized due to
weather, low market demand, or failures in storage, transportation, or processing. The term "food waste" generally
refers to food that is not consumed due to inefficiency or choice at the retail and consumption stage. In this report
the term "consumption" (or the "consumption" stage of the food supply chain) is used to denote the receipt of food
by consumers for use at home or away from home. This term is used regardless of whether the food is ultimately
eaten (i.e., it is not used to mean the biological ingestion of food).
1.4 Report Overview
The report begins by examining the resource requirements and environmental impacts of the U.S. food system
(Chapter 2) and characterizing the amount, type, and timing (i.e., at which stages of the supply chain) of FLW
currently in the U.S. (Chapter 3). Chapter 4 marries these data to estimate the cradle-to-consumer environmental
footprint of U.S. FLW. Chapter 5 summarizes the potential environmental benefits of halving FLW in the U.S., and
Chapter 6 provides global context, highlighting key similarities and differences between the U.S. and global
averages and the U.S. and other developed nations. Chapter 7 summarizes the findings of this report and
identifies key research needed to help the U.S. achieve its 2030 Food Loss and Waste Reduction Goal with
maximum environmental benefits.
Chapter 1. Introduction
3
-------
1.5 Methods
Preparation of this report began with a systematic literature search to identify and collect relevant peer-reviewed
publications, book chapters, and other publicly available information pertinent to FLW. The literature search
included references from 2010 through 2020, with priority given to publications from 2014 or later. Additional,
more recent sources were added during the review process. Appendix B provides further details about the
literature search methods, including key words, literature databases, and screening methods.
Most of the sources cited in this report are peer-reviewed publications. This report also references a number of
government and intergovernmental reports and data sources, which may or may not have been peer reviewed.
Examples include food supply and demand data from the U.S. Department of Agriculture (USDA) and the Centers
for Disease Control and Prevention (CDC); food loss and waste estimates from EPA, USDA, and CEC; and
information about global food loss and waste from the Food and Agriculture Organization (FAO) of the UN.
Additional gray literature from non-governmental organizations is referenced to provide context, such as the
commonly cited food loss and waste estimates from ReFED and the Natural Resources Defense Council (NRDC)
and additional information from the World Resources Institute (WRI) and the UK Waste and Resources Action
Programme (WRAP).
Chapter 1. Introduction
4
-------
CHAPTER 2.
Environmental Footprint of the
U.S. Food Supply Chain
The U.S. food system requires heavy use of the nation's land, water,
and other finite resources, the use of which directly and indirectly
affects environmental quality . This chapter provides an overview of
the environmental footprint of the U.S. food supply chain, from cradle
to consumer, as a foundation for understanding the environmental
footprint of U.S. FLW presented in Chapter 4.
2.1 U.S. Food Supply Chain
The U.S. cradle-to-consumer food supply chain starts with
agriculture and ends in the hands of consumers. While the definitions
and organization of stages vary in the literature, the major stages of
the U.S. cradle-to-consumer food supply chain typically include:
1. Primary production: Farming and harvesting of
plants and animals, resulting in raw food materials.
2. Distribution and processing: Packaging,
processing, manufacturing, transporting, distributing,
and wholesale vending of food and food products.
3. Retail: Selling food and food products to the public
at supermarkets or other stores.
4. Consumption: Receiving food at home or away
from home, such as at restaurants, cafeterias,
institutions, or other locations, regardless of whether
the food is ultimately eaten or wasted.
The first two stages (primary production and distribution and
processing) are often referred to as "upstream" in the supply chain,
while the latter two stages (retail and consumption) are referred to as
"downstream" from the earlier stages. As discussed in Chapter 1,
while the management of FLW (e.g., via landfills, combustion,
composting, or anaerobic digestion) is a part of the food system, it is
not addressed in this report.
KEY FINDINGS
¦ The U.S. cradle-to-
consumer food supply
chain includes four stages:
(1) primary production, (2)
distribution and
processing, (3) retail, and
(4) consumption (food
service and households).
¦ Of the four stages, primary
production accounts for
most of the land, fertilizer,
and pesticide use, plus the
largest share of blue water
withdrawals and GHG
emissions. The
consumption stage uses
the largest share of
energy.
¦ Among food categories,
animal products require
the most land, water,
fertilizer, and energy and
emit the most GHGs per
unit of food.
- Approximately one-fifth of
the U.S. food supply is
imported; however, most
studies assume all food is
produced domestically
and do not account for
differences in the
environmental footprint of
production in different
areas, including water
scarcity, deforestation,
and other factors that lead
to biodiversity loss.
Chapter 2. Environmental Footprint of the U.S. Food System
5
-------
2.2 Environmental Footprint
A wide variety of resources - including land, water, energy, and chemical inputs - are used by the U.S. cradle-to-
consumer food supply chain. Figure 2-1 provides a snapshot of the environmental footprint based upon estimates
from Canning et al. (2020)2 and Crippa et al. (2021 )3. The figure depicts a simplified model of five major inputs
and greenhouse gas (GHG) emissions associated with each stage of the food supply chain and denotes
percentage contributions, by supply chain stage.
As shown in Figure 2-1, primary production is responsible for the widest range of environmental inputs among the
stages of the U.S. cradle-to-consumer food supply chain. The use of land and the application of pesticides and
fertilizers occur chiefly during primary production, while the use of water and energy and the emissions of
greenhouse gases occur all along the food supply chain (Crippa et al., 2021; Canning et al., 2020).
INPUTS
SUPPLY
CHAIN
IMPACTS
GHG Emissions
Non-C02 ¦¦¦¦
co, wmm
FIGURE 2-1. ENVIRONMENTAL FOOTPRINT OF THE U.S. CRADLE-TO-CONSUMER FOOD SUPPLY CHAIN
This figure portrays the use of five major inputs and the emission of greenhouse gases, by supply chain stage.
Data Source: Canning et al. (2020); Crippa et al. (2021)
2 Canning et al. (2020) combined three models (a diet model, an environmentally extended input-output model of resource use
in the food system, and a biophysical model of land use for crops and livestock) to estimate resource use. The study examines
only domestic production, excluding resources used to produce exports and resources used in other countries to produce food
imported into the United States.
3 Crippa et al. (2021) developed a database of global food emissions (EDGAR-FOOD), including emissions associated with
land use and land-use changes, from the existing Emissions Database of Global Atmospheric Research (EDGAR). The
database extends from 1990-2015 and covers all stages of the food chain for every country. With this data the authors
analyzed global food-system emissions and trends and evaluated key contributors (by supply chain stage and country).
LAND
100%
100%
PESTICIDES
WATER
ENERGY
100%
66% 5% 4% 24% 17% 32%10%
41%
PRIMARY
PRODUCTION
DISTRIBUTION
& PROCESSING
RETA L
CONSUMPTION
32% 13% 87%
Chapter 2. Environmental Footprint of the U.S. Food System
6
-------
2.3 Inputs and Environmental Impacts
This report focuses primarily on five inputs to the U.S. cradle-to-consumer food supply chain—agricultural land
use, water use, application of pesticides and fertilizers, and energy use—plus one environmental impact—
greenhouse gas emissions. This section discusses these factors and describes their connection to the major
environmental impacts of the cradle-to-consumer food supply chain, such as climate change and reductions in
biodiversity and water quality and availability. Additional information on inputs and environmental impacts is
provided in Appendix A via reference tables sorted by broad food category (plants, farm animals, and wild-caught
and farmed seafood) and stage of the supply chain. These tables include information about inputs and
environmental impacts beyond those detailed in the report, such as soil degradation, changes in air quality, and
worker health.
Agricultural Land Use
Land is a limited resource integral to the production of food. This report focuses on the land used to produce food
for human consumption, including land used to house livestock and produce feed (e.g., hay, feed grains, and
oilseeds) for livestock and farmed seafood, where possible. Land used for non-food crops and biofuel feedstocks
is excluded.
Over 25 percent of the United States' total land area is used to produce food, with annual per person estimates
ranging from 3,399 to 10,800 m2 (Canning et al., 2020; Birney et al., 2017; Peters et al., 2016). When land is used
to produce food, it can alter soil, air, and water quality (Aneja et al., 2009; U.S. EPA, 2005; USDA, 1996a, b, c).
Of the land required for U.S. food production, 830,000 km2 are used
as cropland and more than 3 million km2 are used as pasture (i.e.,
for grazing). While cropland represents a smaller share of total land
use, it generally requires greater inputs (i.e., fertilizer, pesticides)
and cultivation (e.g., tillage) than pasture, leading to greater
environmental impacts. Of the cropland, approximately 530,000
km2 (63 percent) are used to grow feed for livestock (Merrill and
Leatherby, 2018). Due to feed and pasture requirements, animal
products have a larger land footprint per kilogram than plant-based
foods, with beef requiring significantly more land than other animal
products (Bozeman et al., 2019; Eshel et al., 2014).
While the amount of land used for agriculture in the United States
has been fairly stable since before the 1960s (USDA, 2017), natural
ecosystems are being converted to agricultural land in other
countries (in some cases, to meet demand in the United States -
see Section 2.4), and researchers project a need to further expand
agricultural land to feed the growing global population (UN, 2020a;
Searchinger et al., 2019). Expanding agricultural land use can lead
to loss of biodiversity and ecosystem services (e.g., pollinators, soil
fertility, water filtration) and release of greenhouse gas emissions,
as well as affect hydrologic cycle and local climates (Chaudhary
and Kastner, 2016; Power, 2010; Foley et al., 2005).
Water Use
The food system depends on freshwater for many functions, from
irrigating crops to processing food products to preparing and
cooking food. Like land, usable freshwater supplies are limited, and
many parts of the United States have already reached a "water-
stressed" state (Marston et al., 2021; Capel et al., 2018). As shown
in Figure 2-1, primary production is a major user of freshwater.
Producing plants, animals and farmed seafood for human
consumption means there is less water available for other uses
(Pfister et al., 2011; Rost et al., 2008).
Calculating the
Environmental Footprint
of Animal Products
When estimating the environmental
footprint of producing meat and other
animal products (e.g., dairy or farmed
seafood), researchers should account
for the inputs and impacts associated
with feeding and housing the animals.
For example, the environmental
footprint of pork production includes not
only the resources associated with
housing the animal (e.g., land, water
and electricity use for housing the
animals and GHG emissions from
manure management), but also the
resources required for growing animal
feed (e.g., the land, water, fertilizers,
pesticides and energy required to grow
corn, oats, soybeans and other animal
feed and the GHG emissions
associated with their production).
Where possible, these comprehensive
estimates of the inputs and impacts of
animal products are included
throughout the report.
Chapter 2. Environmental Footprint of the U.S. Food System
7
-------
In addition, water stress can affect aquatic organisms and groundwater-dependent terrestrial ecosystems (Pfister
et al., 2011). With climate change altering hydrologic patterns, water scarcity will likely increase in the United
States and globally (Distefano and Kelly, 2017).
In this report, water use is classified by its source and/or purpose:
1. Blue water - freshwater from surface water and groundwater;
2. Green water - rainwater that is soaked up by cropland or rangeland vegetation or soil; and
3. Grey water - freshwater that is needed to dilute pollutants to meet water quality standards
(Mekonnen and Hoekstra, 2011; Rost et al., 2008).
This report focuses primarily on blue water, with green water data provided when available. Unless otherwise
noted, the term water in this report refers to blue water. Gray water is not consumptive (in that plants, livestock or
farmed seafood do not consume grey water like they do blue and green water), and it is difficult to measure
(Mekonnen and Hoekstra, 2011), and thus is excluded from this report.
All stages of the U.S. cradle-to-consumer food supply chain require blue water. In total, the cradle-to-consumer
food supply chain is responsible for approximately 30 percent of U.S. blue water withdrawals (approximately 34
trillion gallons annually) (Rehkamp and Canning, 2018). Primary production accounts for the largest volume of
water withdrawals (Canning et al., 2020), principally for irrigation (63 percent) (USDA, 2021b). Aquaculture (i.e.,
farming of aquatic organisms) (2 percent) and livestock watering and hygiene (less than 1 percent) account for
smaller shares of blue water withdrawals (Dieter et al., 2018). However, measuring freshwater withdrawals (i.e.,
how much water is withdrawn from surface or groundwater) during primary production may overestimate the
amount of water agriculture consumes4, as only approximately half of the water withdrawn is taken up by plants
and the other half recharges groundwater or soils (Bhagwat, 2019). Much of the water used in aquaculture flows
through the farm and is returned after use (USGS, 2019). To measure water consumption, models are used to
determine crop requirements and irrigation efficiencies. For example, using these methods, it is estimated that 62
percent of irrigation water is consumed (Dieter et al., 2018). Of consumptive blue water use, Marston et al. (2018)
estimates 56 percent is from groundwater, and 44 percent is from surface water sources. In this report "water
withdrawal" or "water use" data and "water consumption" data are distinguished where possible.
During primary production, the amount of blue water required per kilogram of food produced varies widely by food
category. About 80 percent of vegetable crops and 94 percent of orchard fruit and nut crops are irrigated in the
United States (USDA, 2015b, 2013), however meat and other animal products can require a larger amount of blue
water (per unit of food) once irrigation of animal feed is included (Bozeman et al., 2019; D'Odorico et al., 2018;
Eshel et al., 2014).
Downstream from primary production, blue water is utilized during food processing (as a food ingredient and for
processing operations, cleaning, and sanitation) and during the consumption stage (for food preparation and
cooking) (Bhagwat, 2019). Not surprisingly, beverages account for the greatest blue water withdrawals during
food processing, followed by processed foods and flavorings and livestock slaughtering and processing (Marston
et al., 2018). However, total food processing blue water withdrawals account for a very small fraction of cradle-to-
consumer food system water withdrawals, which are dominated by the primary production and consumption
stages (Canning et al., 2020; Marston et al., 2018).
Green water is also critically important for primary production. In the United States, green water comprises nearly
87 percent of consumptive water use to grow crops (Marston et al., 2018). In 2018, only 8 percent of U.S.
cropland and grazing land was irrigated, meaning that 92 percent of crop and grazing land depended solely on
green water for successful production (USDA, 2019b, e). By food category, meat, poultry and eggs, followed by
dairy, are the largest users of green water in the current U.S. diet (Birney et al., 2017). When both blue and green
water are considered, primary production accounts more than 95 percent of the total consumptive water use of all
U.S. economic production (Marston et al., 2018). While the data presented in this report is at national scale, water
scarcity (also called water stress) typically occurs at a smaller scale. However, a regional analysis of water supply
and demand was beyond the scope of this paper.
4 In this context, "consumed" means taken up by the crop over its various growth stages for plant retention and
evapotranspiration.
Chapter 2. Environmental Footprint of the U.S. Food System
-------
Pesticide and Fertilizer Application
Farmers use pesticides on their fields or pastures to protect against yield loss or damage (USDA, 2019b).
Pesticides are substances or mixtures of substances intended for preventing, destroying, repelling, or mitigating
plant or animal pests. Based on the type of crop grown and the pest(s) of concern, farmers may apply natural or
synthetic herbicides, insecticides, fungicides, soil fumigants, plant growth regulators, defoliants, and/or desiccants
to control pests (Hellerstein et al., 2019). Herbicides are the most applied pesticide (63 percent) in the United
States, followed by sulfur and oil and fumigants (Atwood and Paisley-Jones, 2017). Corn (40 percent), soybeans
(22 percent), and potatoes (10 percent) account for the greatest share of pounds of pesticides applied
(Fernandez-Cornejo, 2014; Aktaret al., 2009). The application of pesticides can contaminate waterways, impact
soil quality, and cause harm to ecosystems, and pose risks to non-target organisms including humans (Capel et
al., 2018; U.S. EPA, 2005).
Synthetic and organic fertilizers increase crop yields through the addition of essential plant nutrients. The three
major types of synthetic fertilizer used in the United States are nitrogen (N), phosphate (P), and potassium or
potash (K). The type and amount of fertilizer applied to each crop varies based on local soil conditions, farm
practices, and individual crop needs. In 2015, a total of 20 million metric tons of fertilizer were applied in the
United States, comprised of 11.8 million metric tons of nitrogen, 3.9 million metric tons of phosphorous, and 4.3
million metric tons of potassium (USDA, 2019a). More than half of the fertilizer was applied to feed crops (i.e., to
produce animal products), while the remainder was used primarily on grains and sweeteners (Toth and Dou,
2016).
Nitrogen (N) is found primarily in an organic form in soils but can also occur as nitrate, which is extremely soluble
and mobile. Phosphorus (P) occurs in soil in several forms, both organic and inorganic. Phosphorus loss due to
erosion is common, and phosphate, while less soluble than nitrate, can easily be transported in runoff. Potash is
the oxide form of potassium (K); its principal forms as fertilizer are potassium chloride, potassium sulfate, and
potassium nitrate. When used at recommended application rates, there are few to no adverse effects from
potassium, but it is a common component of mixed fertilizers used for high crop yields (Weil, 2017).
Application of fertilizer can lead to adverse environmental impacts. Nutrient run off from nitrogen and
phosphorous fertilizers can result in drinking water toxicity, eutrophication of streams, and algal blooms and fish
kills (USGS, 2000). Additionally, the production of synthetic fertilizer and the application of organic and synthetic
fertilizers produce greenhouse gas emissions which contribute to climate change (U.S. EPA, 2019b).
Energy Use
All stages of the cradle-to-consumer food supply chain require energy (i.e., electricity and/or fuel). Energy is used
for everything from fueling tractors and pumping and distributing irrigation water to running food processing
equipment to powering refrigeration. Between 2004 and 2015, the U.S. cradle-to-consumer food supply chain
used an average of 11,800 PJ annually, equivalent to 11 percent of total U.S. energy use (Pagani et al., 2020;
Vittuari et al., 2020).
Unlike with the inputs discussed previously, food processing (including packaging) is a significant energy user,
accounting for roughly one-quarter of the cradle-to-consumer food system's energy use (Canning et al., 2020).
Retail accounts for another quarter of energy use. The consumption stage accounts for more than one third of
energy use (the largest share), primarily from refrigeration and cooking (Canning et al., 2020). Despite the level of
attention it receives, transportation from farm to retail (or food service)5 accounts for only approximately 6 percent
of cradle-to-consumer food supply chain energy use (Pagani et al., 2020; Vittuari et al., 2020). By food category,
energy use is highest for meat and dairy, followed by grains (Pagani et al., 2020; Vittuari et al., 2020).
5 This estimate excludes transportation from retail to homes.
Chapter 2. Environmental Footprint of the U.S. Food System ^¦
9
-------
Greenhouse Gas Emissions and Climate Change
Greenhouse gases, including carbon dioxide, methane, nitrous oxide, and some synthetic chemicals, including
chlorofluorocarbons, trap some of the Earth's outgoing energy, thus retaining heat in the atmosphere. Human
activities are increasing the concentrations of GHGs in the atmosphere and are the primary cause of the 1 degree
Celsius increase in global air surface temperature over the past 115 years (Wuebbles et al., 2017). The effects of
climate change on global natural systems include increases in land, water and air temperatures, variation in
precipitation timing and amounts, reduced snow pack, sea level rise, and wildfires and hurricanes (Dupigny-
Giroux et al., 2018; Wuebbles et al., 2017). The Paris Agreement set global targets to limit warming below 2
degrees Celsius, with aspirations to keep warming to 1.5 degrees Celsius because warming beyond the 1.5
degrees Celsius target would lead to more catastrophic outcomes (IPCC, 2018; UNFCC, 2015). Multiple studies
have concluded that reducing GHG emissions from our food system will be essential to feed the growing global
population sustainably and keep food-related emissions in line with limiting global warming to below 2 degrees
Celsius (Clark et al., 2020; Willett et al., 2019; Conijn et al., 2018; Springmann et al., 2018; Bajzelj et al., 2014).
As shown in Figure 2-1, GHGs are emitted at all stages of the U.S. cradle-to-consumer food supply chain, with
the amount and type varying by stage. Studies agree that the greatest amount of GHG emissions occur during
primary production (Crippa et al., 2021; Canning et al., 2020; Boehm et al., 2018 ; Mohareb et al., 2018; Weber
and Matthews, 2008). Examples include:
Methane (CH4) emitted from enteric fermentation6, manure management and growing rice;
Nitrous oxide (N2O) emitted from nitrogen fertilization and manure management; and
Carbon dioxide (CO2) emitted from soil management practices (i.e., reduction in soil carbon
sequestration resulting in release of CO2 into the atmosphere), fertilizer production, and energy
use by farm equipment.
Primary production in the United States releases approximately 4.72 kg CC>2e (CO2 equivalents) per person per
day (Heller et al., 2018) and is responsible for 39 percent of U.S. methane emissions and 80 percent of U.S.
nitrous oxide emissions (U.S. EPA, 2021c). Both methane and nitrous oxide are potent greenhouse gases, with
global warming potentials more than 25 and 265 times greater than CO2 (U.S. EPA, 2021c). Methane has only a
short (12-year) atmospheric life. Nitrous oxide is also the most significant ozone-depleting substance released to
the atmosphere, damaging the stratospheric ozone layer that protects Earth from the sun's harmful radiation
(Compton, 2021). Globally, land clearing and deforestation is also a major source of GHG emissions.
As shown in Figure 2-1, roughly half of the cradle-to-consumer food supply chain's GHG footprint is CO2
emissions from energy use and land use change. Energy use drives the GHG emissions of all of the supply chain
stages downstream from primary production, with the exception of retail, which emits chlorofluorocarbons (CFCs)
from refrigerant leaks (Crippa et al., 2021) in addition to CO2 from energy use. Transportation contributes a
relatively small share of food system GHG emissions, representing from 7 to 11 percent of U.S. cradle-to-
consumer food supply chain emissions, according to recent studies (Mohareb et al., 2018; Weber and Matthews,
2008).
Many studies have examined GHG emissions by food category. Despite differences in methodologies, portions of
the food system covered, and other variables, most studies found that the production of meat (especially beef)
results in the most GHG emissions per weight of food produced (Guo et al., 2020; Bozeman et al., 2019;
D'Odorico et al., 2018; Birney et al., 2017; Heller and Keoleian, 2015; Venkat, 2012).
6 Enteric fermentation is fermentation that takes place in the digestive systems of animals, resulting in methane that is exhaled
or belched by animals.
Chapter 2. Environmental Footprint of the U.S. Food System ^¦ 10
-------
Summary
In summary, the U.S. cradle-to-consumer food system is significant user of finite natural resources and
contributes to a broad range of environmental impacts, including climate change. The inputs and environmental
impacts of the food supply chain vary by both supply chain stage and by the category of food being produced.
The primary production stage of the supply chain is responsible for most land, pesticide, and fertilizer use, and the
greatest share of water withdrawals and GHG emissions. Energy use, however, is dominated by the consumption
stage, followed by the food processing and packaging stage, thus these stages also contribute significantly to
GHG emissions and climate change (Canning et al., 2020). In general, the production of animal products requires
the greatest amount of land, water, and energy and results in the most GHG emissions per weight of food
produced (Pagani et al., 2020; Vittuari et al., 2020; Bozeman et al., 2019; D'Odorico et al., 2018; Hilborn et al.,
2018; Parker et al., 2018; Aleksandrowicz et al., 2016; Tom et al., 2016; Weber and Matthews, 2008).
2.4 Imports
U.S. consumers often rely on imported agricultural goods when domestic production is not possible (e.g., grapes
during the U.S. winter season), when demand outweighs domestic production capacity, or for other reasons. For
example, U.S. production of fruits and vegetables such as bananas and asparagus is able to meet less than 20
percent of domestic demand, whereas for other produce (e.g., apples, oranges, and cauliflower) the United States
produces more than is demanded domestically and is a net exporter (FAO, 2018).
According to the USDA, almost one-fifth of the food consumed within the U.S. is imported. Fruits (26 percent) and
vegetables (20 percent) constitute almost half (46 percent) of all agricultural imports combined, and imports
supplied approximately 30 percent of the available vegetables and more than half of all the available fruits in the
U.S. (USDA, 2019f, 2016). Fish and seafood are also frequently imported, with imports comprising over 80
percent of the seafood available to U.S. consumers (NOAA, 2021).
Geographic Differences in Environmental Footprint
Figure 2-2 shows U.S. food imports by country in 2019. While these imports come from diverse locations, most of
the studies of the environmental footprint of FLW assume the entire U.S. food supply is produced domestically
when calculating environmental footprint. This is due to the complexity of attributing imports to producer countries
and estimating environmental impacts of each food in each producing country. The studies typically apply
average U.S. resource use levels and emission factors to all food available to U.S. consumers, regardless of
whether it was produced in the United States. Thus, the studies may over- or under-estimate the actual
environmental footprint of production, since the environmental footprint of producing food in another country may
vary greatly from that of producing food in the United States.
Differences in producing countries' local environments, agricultural practices, yields, standards, and production
methods can all impact the environmental footprint of a food item (Poore and Nemecek, 2018). For example, the
production of imported goods can contribute to water scarcity or agricultural land use change and deforestation in
exporting countries. While the level of land use for agriculture is relatively stable in the United States (USDA,
2017), natural ecosystems may be converted to produce crops or graze livestock in other countries. A recent
study by Kim et al. (2020) compared the GHG footprint and blue and green water footprints of producing 74 food
items in different countries and found substantial variation in impacts of food production by country. The authors
found, for example, that the GHG footprint of beef produced in Australia (the top importer of beef to the U.S.) was
nearly double that of beef produced in the United States, while the blue water footprints of rice produced in India
and Thailand (leading importers of rice to the United States) were half that of rice produced domestically (Kim et
al., 2020; USDA, 2020).
A few FLW studies presented in this report attempt to account for differences among producing countries. One
study (Jalava et al., 2016) applied national average environmental factors to domestically produced food and
global average environmental factors to imported food, while two others (Guo et al., 2020; Chen, 2019) applied
global average environmental factors to all food (domestically produced and imported). A fourth study (Pagani et
al., 2020; Vittuari et al., 2020) simply excluded imports from their analysis, thus underestimating the
environmental footprint of FLW by not including the footprint of producing 20 percent of the U.S. food supply. No
food waste studies applied country-specific data to imports when estimating environmental impacts of FLW.
Chapter 2. Environmental Footprint of the U.S. Food System
11
-------
U.S. Food Supply - by value in 2019
Domestic
Imported
£11
US?
¦iT
Value of Annual U.S. Imports
More than $26B
~ $3.5B to $26B
~ $2B to $3.5B
~ $1.2B to $2B
FIGURE 2-2. IMPORTS TO U.S. FOOD SUPPLY
Food imports to the United States are predominantly from Mexico and Canada, followed
by France, Italy, Chile, China, India, and Indonesia, in order of value of imported food.
Groupings in figure correspond to the 99th, 95th, 90th, and 85th percentiles across 201 importing countries.
Data Source: CRS (2020); World Bank (2021)
Biodiversity Loss
In the studies presented in this report, agricultural land use is often used as an indicator of the potential for
biodiversity loss, and many examples exist of species threats due to global trade. A systematic analysis linking
threatened species, commodities, and complex international supply chains through 187 countries by Lenzen et al.
(2012) showed that 30 percent of global species threats7 are due to international trade. The study also concluded
that the United States is the largest net exporter of biodiversity threats, meaning that primary production of foods
and other goods imported into the U.S. (including coffee, tea, sugar, textiles, fish and other manufactured items)
threaten the greatest number of species abroad.
7 Excluding threats from invasive species..
Chapter 2. Environmental Footprint of the U.S. Food System
12
-------
Looking specifically at the biodiversity impacts of food imports, Chaudhary and Kastner (2016) used the
countryside species area relationship (SAR) model, paired with bilateral trade data from FAO, to identify species
lost due to agricultural land use for 170 crops in 184 countries. The model estimated that U.S. food imports are
responsible for the loss of 115 species abroad, primarily in Mexico and Indonesia, and also in other Latin
American countries (Chaudhary and Kastner, 2016). An additional study using the SAR model identified the
export of agriculture-based products from Mexico and the Philippines and pasture-based products from Australia
and Columbia, along with forestry-based products from Oceania, Indonesia, the Philippines, and Central Africa,
as the greatest biodiversity impacts associated with global trade (Chaudhary and Brooks, 2019). In addition, some
food products such as sugarcane, palm oil, and coffee have disproportionately high impacts on biodiversity
relative to the amount of land occupied by their production (Chaudhary and Kastner, 2016); thus, the studies
presented in this report may miss important impacts on biodiversity due to the use of broad categories of foods.
International Transportation
As imported foods require transportation from their country of origin to the United States, additional environmental
impacts may occur from shipping and storage of food during transportation (e.g., refrigeration for perishable
animal products or fruits, or freezer conditions for frozen fish or vegetables). Note, however, that depending on
where food is produced and consumed, imported food could travel less distance than domestically grown food
(e.g., wheat imported from western Canada to Minnesota compared with oranges grown in Florida and sold in
Hawaii). Impacts from travel also vary considerably by method, with longer distance travel by boat sometimes
resulting in lower GHG emissions than shorter distance travel by truck (Wakeland et al., 2012). Also, differences
in GHG emissions related to production methods can outweigh those associated with transportation. Only one
study of FLW presented in this report (Guo et al., 2020) accounted for international transportation when
estimating GHG emissions. The authors found that it was equivalent to just 3 percent of the GHG emissions from
primary production (measured in CO2 equivalents).
2.5 Other Factors
Studies estimating the inputs and environmental impacts of each type of food (and FLW) at each stage of the U.S.
food system utilize national or regional averages for inputs and environmental impacts, typically for broad
categories of foods, such as all fruits or all grains. In reality, inputs and impacts vary depending on multiple
factors, including:
Type of food produced (within a broad category of food),
Production method,
Geographic location and timing of production,
Type and amount of processing conducted (e.g., no processing for corn on the cob versus
processing for corn meal or extensive processing for high-fructose corn syrup),
Type and amount of packaging used,
Type of storage required (e.g., refrigeration),
Time between being produced and purchased (seasonality),
Mode and distance of transportation at each stage,
How the food is cooked, stored and prepared, and
Other factors (Asem-Hiablie et al., 2019; Bozeman et al., 2019; Goossens et al., 2019; Heard et
al., 2019; Niles et al., 2018; Clark and Tilman, 2017).
Each of these variables can affect the cumulative resource requirements of the finished food product as well as
the type, amount, and cumulative environmental impacts of its production and distribution (Asem-Hiablie et al.,
2019; Heard et al., 2019; Niles et al., 2018) - including all associated FLW. The studies presented in Chapters 4
through 6 of this report estimate the approximate average environmental impacts at each stage of the cradle-to-
consumer food supply chain for each category of food lost or wasted. This means the studies do not calculate the
precise environmental impacts accrued at each stage of the food system for specific FLW (e.g., the leftover steak,
potatoes, and green beans one threw away in a household), but instead typically calculate the environmental
impacts of the broad categories of FLW (e.g., meat and vegetables) during broad stages of the food supply chain
(e.g., the consumption stage, which includes at home and away from home consumption).
Chapter 2. Environmental Footprint of the U.S. Food System
13
-------
CHAPTER 3.
Characterization of
U.S. Food Loss and Waste
The American food system is a complex arrangement of farmers,
processors, distributors, retailers, food service providers, and
consumers. Key players can influence what is produced and how
products move through the system (i.e., through business decisions
and consumer preferences), but the system as a whole is driven by a
multitude of factors, including domestic and global markets, costs,
politics, laws and regulations, social organizations, plant and animal
biology, science and technology, weather, and environmental
conditions (IOM and NRC, 2015). Combined, these factors determine
the total amount and types of food produced, consumed, lost, and
wasted each year.
Understanding the amount of food produced for human consumption
but ultimately lost or wasted is an important step toward assessing the
magnitude of the environmental footprint of FLW. This chapter
examines published estimates of the total amount of FLW generated
in the United States along with details regarding the categories of food
lost or wasted and the supply chain stage at which food is lost or
wasted. This information is critical to building the estimates of the
environmental footprint of U.S. FLW presented in Chapter 4 and to
tailoring efforts to reduce food loss and waste.
3.1 U.S. Food Surplus
America has an overabundance of food. According to the USDA, the
amount of food available to U.S. consumers is far greater than the
amount of food they consume. As shown in Figure 3-1, 3,796 to 4,000
calories were available8 per person per day, compared to 2,081
calories consumed per person per day, in 2010 (USDA, 2019d,
2015a; Buzby et al., 2014; USDA, 2012).9 This indicates that
significant FLW is an outcome of the U.S. food system.
Wasted food also represents wasted nutrients, which vary by food
category wasted. Spiker et. al (2017) found that food wasted by
retailers and consumers in 2012 contained 33 grams protein, 5.9
grams dietary fiber, 1.7 micrograms vitamin D, 286 milligrams
calcium, and 880 milligrams potassium per person per day (Spiker et
al., 2017).
KEY FINDINGS
¦ The U. S. wastes more
than one third of its food
supply, from 73 to 152
million metric tons (161 to
335 billion pounds) per
year or 223 to 468 kg
(492 to 1,032 pounds)
per person per year
• U.S. FLW includes 1,110
to 1,520 calories per
person per day.
¦ U. S. FL W per person
increased over the last
decade and total U. S.
FLW tripled since1960.
¦ The consumption stage
(restaurants and
households) is
responsible for roughly
half of U.S. FLW.
¦ Fruits and vegetables are
the most commonly
wasted foods, followed by
dairy and eggs.
8 Data on food availability come from USDA's Food Availability Per Capita Data Series (Buzby et al., 2014). 2010 is the most
recent year for which complete data is available. The FAO provides an updated estimate of 3,782 calories per person per day
in 2018 (FAO Food Balances, 2021).
9 Data on consumption come from USDA's National Health and Nutrition Survey, What We Eat in America (2009-2010 data)
(USDA, 2012). More recent estimates (2017-2018) from this data source indicate 2,093 calories consumed per person per
day.
Chapter 3. Characterization of U.S. Food Waste
14
-------
While food is abundant in the U.S., food insecurity persists, in 2019, more than 35 million Americans were food
insecure (USDA, 2021 a). However, this food insecurity is not driven by scarcity. As shown in Figure 3-1, studies
indicate that even if every American was provided with enough calories to meet their current level of physical
activity and body weight, a surplus of 1,050 to 1,400 calories daily per person would remain (Hip et ai., 2016; Hail
et al., 2009)10. The amount of surplus food from retailers and consumers (141 trillion calories in 2010, according
to Buzby et al. (2014)) is sufficient to feed 154 million people for a year (Wood et al., 2019), a far greater number
than estimated by USDA to be food insecure. In addition, it is not always possible or appropriate to redistribute
surplus food (Spikeret al., 2017). Therefore, increasing the redistribution of food cannot alone meet the U.S. goal
to halve food waste by 2030. Solutions must include efforts to prevent the generation of surplus food and FLW in
addition to efforts to redistribute surplus food where possible.
Food Available to Americans
Surplus Food
Food Consumed
Enough for all
food insecure
Americans
Remaining Food Waste
FIGURE 3-1. FOOD WASTE IN THE UNITED STATES
The amount of food available to Americans (in calories) exceeds the number of calories consumed
plus the number of calories required to eliminate food insecurity. The figure depicts only
edible food (i.e., inedible parts such as bones and shells are excluded from estimates). Data year 2010.
Data Source: Buzby etal. (2014); USDA (2012); Hall etal. (2009); Hig et al. (2016)
10Surplus calories are estimated using biological models for human energy requirements and loss adjusted food availability estimates. More information on the
methods used to develop these estimates is available in section 3.7.
Chapter 3. Characterization of U.S. Food Waste
15
-------
3.2 Total U.S. FLW
Given the size and dynamic complexity of the U.S. food system, no single agreed-upon comprehensive estimate
of the total amount of U.S. FLW exists. Instead, the literature includes multiple credible estimates, which differ in
scope and methodology, that together provide insights into the magnitude and distribution of U.S. FLW.
Table 3-1 provides a summary of the estimates of total U.S. FLW from the literature. All of the estimates include
only food intended for human consumption. Three key variables that impact the magnitude of these estimates -
data year, edibility, and supply chain coverage - are shown, and their effect will be discussed in the following
three sections. Note that while there a large number of FLW estimates, many rely upon a similar data sources
(most notably, data series from FAO and USDA). See Section 3.7 for a discussion of these methodologies.
Of the studies that include all FLW11 from all stages of the food supply chain (from primary production to
consumption), the estimates of U.S. FLW range from 73 to 152 million metric tons per year, or 223 to 468 kg per
person per year (ReFED, 2021 a; Guo et al., 2020; CEC, 2017). Two of these studies provide results in terms of
percentage of the food supply, equating their estimates of FLW to 35 to 36 percent of the U.S. food supply
(ReFED, 2021a; CEC, 2017).12
11 Include edible and inedible FLW (discussed further in Section 3.4).
12 Equating estimates of U.S. FLW to a percentage of the corresponding food supply can be complex, as each study defines it
boundaries differently. In this report only percentages provided by the study authors are presented.
Chapter 3. Characterization of U.S. Food Waste
16
-------
TABLE 3-1. ESTIMATES OF U.S. FOOD LOSS AND WASTE
U.S. Food Loss and Waste
Source
Scope of FLW
Total
million metric
tons/year
Percent
Food
Supply
Lost or
Wasted
Per Person
By Weight By Calories Data Year
kg/year cal/day
CEC (2017)
33%
368
2007
CEC (2017)
36%
415
2007
FAO (Gustavsson et al., 2011)
300
2007
Guo et al. (2020)
468
2017
Kummu et al. (2012)
i m
Si )^)©)(
32%
1134 2005-2007
Lipinski et al. (2013)
1520
2009
Pagani et ai. (2020) and Vittuari et
al. (2020)b
25%
240
1110 2001-2015
Read et al. (2020)
§4 )© ) (
18% f
by monetary
value
2007-2012
ReFED (2021 a)°
35%
223*
2019
U.S. EPA (2020a)
286*
2018
Hi? eta I. (2016)
1050
2010
NIH (Hall et al., 2009)
30%
1400
2003
Toth and Dou (2016)
45%
309*
2012
Venkat (2012)
180
2009
FAO (2021)
187'
2018
Chapter 3. Characterization of U.S. Food Waste
17
-------
U.S. Food Loss and Waste
Source
Scope of FLW
Total
million metric
tons/yr
Percent
Food
Supply
Lost or
Wasted
Per Person
By Weight By Calories Data Year
kg/year cal/day
Birney et al. (2017)
229
2010
Cuellar and Webber (2010)
¥1
145
2007
Heller and Keoleian (2015)
O ^ 3
31%
196
2010
Lopez Barrera and Hertel (2021)
35%
2013
Mekonnen and Fulton (2018)
o ESS
34%
216
1237
2015
Spiker et al. (2017)
omm
1217
2012
USDA (Buzby etal.,2014)
omm
31%
195'
1249
2010
Chen et al. (2020)
o n n
184'
709
2011
Conrad et al. (2018)
o I s
26%
154
795-840 2007-2014
van den Bos Verma et al. (2020)
XI
1572
2011
Yu and Jaenicke (2020)
HI
only households
32%
by monetary
value
2012
= edible FLW only; • = edible and inedible FLW
Q = Indicates estimate for the North America and Oceania (NAO) Region
• = calculated value. Calculated values for total tons and weight per person used the same population factor as the study's data year.
personal communication with the author. Q. Read (April 5, 2021; July 26, 2021) X. Guo (March 23, 2021); M. Pagani (April 20, 2021;
September 1, 2021)
a Kummu et al. (2012) excludes animal products from FLW estimates.b Pagani et al. (2020) and Vittuari et al. (2020) exclude the waste of
imported foods. Approx. one-fifth of U.S. food supply is imported.c ReFED (2021a) excludes animal products from estimates of FLW during
primary production.
Chapter 3. Characterization of U.S. Food Waste
18
-------
3.3 U.S. FLW Over Time
While trend data are not widely available, ReFED and EPA both provide helpful estimates of U.S. FLW over time.
ReFED's Insights Engine Food Waste Monitor (2021a) provides the most comprehensive data available, including
FLW from all stages of the supply chain.13 Between 2010 and 2019, ReFED estimates that U.S. FLW increased
by 12 percent, including a per person increase of 6 percent. According to ReFED, FLW rates have been relatively
flat since 2016, showing less than a one percent change in both absolute and per person amounts.
U.S. EPA's "Facts and Figures" (2020b) provides data over a longer time period (1960 through 2017) but includes
only FLW from the retail and consumption supply chain stages that is sent to landfills, incinerators, and compost
facilities. Since 1960, EPA reports the amount of FLW in the U.S. has tripled. Between 2010 and 2017, EPA
reports a 14 percent increase in FLW, including a per person increase of 8 percent.
In 2020, EPA revised its food measurement methodology to include additional food waste pathways, FLW
generated by the food processing sector, and more recent studies (U.S. EPA, 2020f). The revised methodology
includes FLW from each sector managed via the three original pathways (landfilling, combustion, and
composting) plus the following six additional pathways: donation, animal feed, anaerobic digestion/co-digestion,
land application, sewer/wastewater, and bio-based materials/biochemical processing.14 This new methodology
substantially increases EPA's estimate of U.S. FLW, as would be expected. Data from both EPA methodologies
are available for the year 2016. In 2016, estimates for FLW from the retail and consumption supply chain stages
from the new methodology (56 million metric tons) are more than 50 percent higher than those calculated using
the former methodology (36 million metric tons) (U.S. EPA, 2020f) due to updated data sources and the addition
of pathways. The addition of the food processing sector accounted for an additional 34 million metric tons; 94
percent of the food processing industry's food waste was managed through the six new pathways added (U.S.
EPA, 2020a). Figure 3-2 portrays trends from ReFED data and both EPA data series.15 All data series show an
increase in FLW over the last decade.
300
250
200
150
100
50
1960
1970
1980
1990
Year
2000
Source
U.S. EPA Revised Methodology (2020a, f)
ReFED (2021a)
U.S. EPA Revised Methodology (2020a, f)
ReFED (2021a)
U.S. EPA Facts & Figures (2020b)
Scope of FLW
2010
2020
FIGURE 3-2. U.S. FLW PER PERSON OVER TIME
Over roughly the last decade, per person food loss and waste in the United States has increased by 6 to 8 percent
(ReFED 2010-2019; U.S. EPA 2010-2017). Over the same time frame, total U.S. food loss and waste increased by 12 to 14 percent.
13 Information on ReFED's methodology can be found in Section 3.7.
14 Data was not available for all pathways for all sectors.
15 Both data series include inedible parts such as bones and shells.
Chapter 3. Characterization of U.S. Food Waste
19
-------
3.4 Edibility of U.S. FLW
One key difference among estimates of FLW is whether they include only edible FLW, or both edible and inedible
FLW, such as bones, pits, and shells. Edibility is based upon the type or part of food, not whether the food was
spoiled when wasted. For example, eggshells are always considered inedible, but the egg inside is always
considered edible, regardless of its age. Estimates of edibility are inevitably rough, since determining and
quantifying edible food parts is not straightforward. The definition of edible food is inherently ambiguous because
what parts of food are intended for human consumption varies along the food supply chain and between individual
consumers and depends on social and cultural preferences and technology factors. For example, the broccoli
stalk is considered edible by some people or cultures but inedible by others. Whether the broccoli stalk is
considered edible or inedible changes the FLW estimate for a head of broccoli by approximately 61 percent
(Moreno, 2020).
Knowing the edible share of FLW (sometimes called avoidable FLW) is important because it provides perspective
on how much of the FLW could have been eaten by people. Edible FLW could go towards feeding people and
lessening food insecurity, whereas inedible FLW is material that, once produced, must be managed (i.e., through
pathways such as composting, anaerobic digestion, or landfill). Knowing both the edible and inedible share of
FLW facilitates resource efficiency and allows policymakers to make more informed decisions on FLW reduction
and management approaches.
Broadly speaking, between 70 to 90 percent of the food lost or wasted in the United States is edible. As shown in
Figure 3-3, the Commission for Environmental Cooperation (CEC) estimated that only 10 percent of FLW (by
weight) from the full food supply chain is inedible (CEC, 2017).16 Estimates of household FLW indicate a greater
percentage of household FLW may be inedible. Kitchen diaries kept as a part of two studies - the Oregon
Department of Environmental Quality's (ORDEQ) Oregon Wasted Food Study17 (McDermott et al.. 2019) and the
Natural Resources Defense Council (NRDC)'s assessment of three cites18 (Hoover and Moreno, 2017) -
demonstrate that inedible FLW accounted for approximately 30 percent of total household FLW, by weight
Inedible
Edible
90%
Supply Chain
Inedible Edible
70%
fil
Households
FIGURE 3-3. SHARE OF U.S. FLW CONSIDERED EDIBLE VERSUS INEDIBLE, BY WEIGHT
Available data indicates that only 10 percent of total U.S. food loss and waste (FLW) and 30 percent of
U.S. household food waste (by weight) are made of inedible food parts, such as bones or shells.
Data Source: CEC (2017); McDermott et al. (2019); Hoover and Moreno (2017)
16 To calculate this estimate, the CEC excluded the conversion factors from the FAO (2011) edible FLW estimates for NAO.
17 Study covered rural and urban areas of Oregon in 2017. 182 households completed 7-day kitchen diaries.
18 Study conducted in Nashville, Denver, and New York City in 2016 and 2017. 613 households completed 7-day kitchen
diaries.
Chapter 3. Characterization of U.S. Food Waste
20
-------
When measuring only edible FLW, estimates of totai FLW are generally lower, as would be expected. The three
studies discussed in Section 3.2 that assessed all U.S. FLW (i.e., edible and inedible) from ail stages of the
supply chain concluded that between 73 to 152 million metric tons, or 223 to 468 kg per person, of FLW per year
(ReFED, 2021a; Guo et al., 2020; CEC, 2017), whereas estimates of edible U.S. FLW from all stages of the
supply chain range from 78 to 112 million metric tons, or 240 to 368 kg per person, FLW per year (Pagani et al.,
2020; Vittuari et ai„ 2020; CEC, 2017).
Examining only edible FLW also allows for an estimation of calories lost or wasted. Figure 3-4 presents estimates
of U.S. FLW, measured by weight and then by calories. Only one estimate of U.S. FLW from all stages of the food
supply chain, by calories, is available. A set of companion studies by Pagani et al. (2020) and Vittuari et al. (2020)
estimate U.S. FLW to be 1,110 calories per person per day. Two additional studies provide comparable estimates
for the North America and Oceania19 (NAO) region, ranging from 1,134 (excluding calories from animal products)
to 1,520 calories per person per day (Lipinski et al., 2013; Kummu et al., 2012).
|®J
|K51
ill
ium
Guo et al. (2020)
CEC (2017)
FAO/Gustavsson
(2011)2
Toth & Dou (2016)
U.S. EPA (2020a)
Pagani et al. &
Vittuari et al. (2020)
ReFED (2021a)
Birney et al. (2017)
Heller & Keoieian
(2015)
Buzbyetal. (2014)
Chen et al. (2020)
Venkat (2012)
Conrad et al. (2018)
Cuellar& Webber
(2010)
1,600
1.400
jo 1.200
c
o
in
i_
d>
a. 1,000
d>
CO
o
800
600
400
Bos Verma et al. (2020)
Lipinski et al. (2013) Q
Birney et al. (2017)
Hall et al. (2009)
Buzby et al. (2014)
Spiker et al. (2017)
Kummu et al. (2012) ^
Pagani et al. (2020) &
Vittuari et al. (2020)
Hig et al. (2016)
Conrad et al. (2018)
Chen et al. (2020)
FIGURE 3-4. U.S. FLW BY ANNUAL WEIGHT AND BY DAILY CALORIES PER PERSON
The figure on the left shows annual estimates of total U.S. FLW by weight,
while the figure on the right shows daily estimates of per person U.S. FLW by calories.
Dotted lines indicate edible FLW only; solid lines represent all FLW, including inedible parts.
The length of each line indicates the stages of the food supply chain included in estimate.
Several studies provide estimates both by weight and by calories, allowing for comparison of changes in rank order.
19 The NAO region, as defined by the FAO, includes the U.S., Canada, Australia, and New Zealand.
Chapter 3. Characterization of U.S. Food Waste
21
-------
3.5 U.S. FLW, by Supply Chain Stage
Understanding when along the supply chain FLW occurs is essential to planning successful food waste reduction
interventions. It is also essential for calculating accurate assessments of the environmental footprint of FLW,
since as food moves through the supply chain, it uses additional inputs and creates additional environmental
impacts. Thus, FLW that occurs further along the supply chain has a larger environmental footprint than similar
amounts and categories of FLW that occur at an earlier stage.
Several studies provide estimates of FLW at each stage of the supply chain. Figure 3-5 shows the relative
contribution to FLW from each supply chain stage, from each of the studies that examined FLW along the entire
food supply chain. As shown in Figure 3-5, studies agree that the greatest share of U.S. FLW occurs during the
consumption stage. The consumption stage accounts for roughly one half of total U.S. FLW. Together, the
consumption and retail stages represent between half and three quarters of aii U.S. FLW (ReFED, 2021a; Pagani
et al„ 2020; Vittuari et al . 2020: CEC. 2017).
11% 7%
18% 7%
21%
13%
13%
53%
CEC (2017)
Pagani et al. (2020) &
Vittuari al. (2020)
ReFED (2021a)
FAO/Gustavsson
et al. (2011)
Guoetal. (2020)
Kummu et al. (2012)
Lipinski et al. (2013)
FIGURE 3-5. SHARE OF U.S. FLW, BY FOOD SUPPLY CHAIN STAGE
This figure illustrates the distribution of food loss and waste (FLW) by supply chain stage from reviewed studies that include all four stages
of the cradle-to-consumer food supply chain. The distribution is by weight of FLW for ail studies except Kummu et al. (2012) and Lipinski et
al. (2013) which are by calories. The consumption stage (households and food service) accounts for roughly one half of U.S. FLW.
33%
21%
7%
39%
35%
17%
13%
35%
5% 6%
15% 7%
Chapter 3. Characterization of U.S. Food Waste
22
-------
As shown in Figure 3-6, weight estimates of U.S. consumption stage FLW range from 34 to 57 miiiion metric tons
per year, equivalent to 110 to 188 kg per person per year. Estimates of combined U.S. consumption and retail
stage waste (i.e., the scope of the U.S. 2030 Food Loss and Waste Reduction Goaf) range from 44 to 71 million
metric tons edible FLW annually, or 143 to 229 kg per person per year.
Food loss during primary production is often left out of FLW estimates, and retail FLW estimates may be
underestimated due to the FLW being attributed upstream or downstream from the retail sector (Read et al.,
2020). Whether and where some FLW is attributed may be inconsistent among data sources, and in many cases
the studies did not provide clear guidelines in this area. For example, when a retailer does not accept a shipment
of produce, the FLW may be attributed to the distribution stage (since the distributor must manage the FLW), or to
the retailer (since the retailer's standards caused the produce to be rejected), or to the consumer (since the
retailer's standards may have been set to meet perceived customer requirements).
H
MJB1
10
10
11
Million Metric Tons
= edible FLW only; • = edible arid inedible FLW
Source
CEC (2017)
FAO/Gustavsson et al. (2011)
Guo et al. (2020)
Pagani et al. & Vittuari et al. (2020)
ReFED (2021a)
Toth&Dou (2016)
U.S. EPA (2020a)
Venkat (2012)
FAO (2019a)
Birney et al. (2017)
Heller & Keoleian (2015)
USDA/Buzby et al. (2014)
Chen etal. (2020)
Conrad et al. (2018)
FIGURE 3-6. AMOUNT OF U.S. FLW, BY SUPPLY CHAIN STAGE, BY WEIGHT
The figure above depicts food loss and waste estimates from al! reviewed studies.
Generally, estimates are highest for the consumption stage (households and food service).
Chapter 3. Characterization of U.S. Food Waste 23
-------
Quantifying FLW during primary production
Of note, the primary production stage (i.e., farming and harvesting of plants and animals) is often partially
or entirely absent from FLW estimates, due to limited data availability. This common exclusion of on-farm
FLW, however, misses a potentially large portion of FLW. The major roadblocks to quantifying losses
during primary production are the difficulty of measurement and the potential for wide variation among and
within food categories, such as fruits and vegetables, seafood, and other animal products.
Studies that include FLW during primary production typically include the loss of fruits and vegetables, but
limited data availability leads to very rough estimates. Fruits and vegetables may be lost in the field for a
variety of reasons - of greatest interest to FLW stakeholders are the losses that may be preventable, such
as those due to produce not meeting grade standards or buying specifications, or due to produce supply
exceeding demand. FLW studies typically begin measuring fruit and vegetable losses once produce is ripe
in the field (Johnson, 2020). Losses due to weather or pests are not included in the FLW estimates
presented in this report.
Data from the Food and Agriculture Organization (FAO) of the United Nations serve as the basis for most
estimates of primary production losses. For all fruits and vegetables, FAO applies a 20 percent loss factor,
but this estimate is not based upon field measurement studies (Johnson, 2020). The limited data available
from U.S. farms shows FAO may underestimate losses of fruits and vegetables, and that wide variation
exists among produce types. For example, a recent field study of nine vegetables on a U.S. farm found an
average field loss rate of 57 percent, including only the produce suitable for harvest and use—much
greater than the 20 percent loss assumed in FAO data (Gustavsson et at., 2013). Another recent field
study demonstrates the wide variation in FLW rates among produce in the U.S., ranging from 2 percent of
potatoes destined for processing, to 56 percent ofromaine lettuce (WWF, 2018).
Seafood losses can be particularly difficult to estimate since more than 80 percent of the seafood available
to U.S. consumers is imported (NOAA, 2021). Bycatch (i.e., non-target aquatic species caught by fishing
gear and discarded dead or injured into the ocean) is a major source of seafood loss. While some FLW
estimates incorporate these losses, the primary method of accounting for these losses is to apply the FAO
loss factor—an average bycatch rate of 12 percent (Gustavsson et al., 2013). However, bycatch rates vary
widely between specific fisheries and species, making estimation difficult (Love et al., 2015), and the FAO
factors is likely an underestimate. Bycatch estimates for the U.S. seafood supply range from 16 to 32
percent in the literature (Love et al., 2015). In addition, none of the FLW estimates presented in this report
incorporate losses due to spoilage of seafood before distribution or losses in aquaculture. NOAA (2021)
estimates that more than half of seafood imported into the U.S. is from aquaculture, making this a
potentially significant source of seafood loss.
Animal products (e.g., meat and poultry) appear to have lower loss rates than produce or seafood (Lipinski,
2020). The FAO (Gustavsson et al., 2013)) estimates only losses for meat, based upon animal deaths prior
to slaughter (less than 5 percent). In summary, only very limited and variable data on primary production
losses is available, leading to exclusion of this potentially important area from FLW estimates.
Chapter 3. Characterization of U.S. Food Waste
24
-------
3.6 U.S. FLW, by Food Category
As described in Chapter 2, the environmental footprint of food (and thus FLW) varies greatly by food category. An
understanding of the categories of food that comprise U.S. FLW is essential to reducing FLW and to estimating its
environmental footprint. When looking at FLW, by food category, across the entire food supply chain, the primary
data source available is FAO (Gustavsson et al., 2011); however, these estimates are for the NAO region, rather
than the U.S. specifically.
When examining FLW from all along the supply chain, the FAO reports that fruits and vegetables are the food
category wasted in the greatest quantity (40 percent of FLW) in the NAO region, as shown in Figure 3-7. Data
from ReFED (2021 a) confirms that produce represents the greatest share (34 percent) of U.S. FLW. Both FAO
and ReFED also note the significance of milk and dairy, and eggs as categories of FLW (20 percent and 16
percent of FLW, respectively).
For many other food categories, the data from ReFED is not directly comparable to that of FAO due to differences
in the way food is categorized. The FAO data (and the USDA data, which is discussed subsequently) categorize
wasted food by commodity ingredients (e.g., fish sticks are classified as fish), whereas ReFED categorizes foods
as grocers do, in their retail form (e.g., fish sticks are classified as frozen foods).
The USDA also provides U.S.-specific data on FLW by food category; however, it is limited to FLW during the
retail and consumption supply chain stages. As shown in Figure 3-7, USDA data (Buzby et al., 2014)
demonstrates that fruits and vegetables (33 percent of FLW) and milk and eggs (15 percent) are the categories
wasted in largest quantities in the retail and consumer stages, consistent with FAO results for all stages in the
data presented above.
North America and Oceania
In some cases, FLW data are not available for the U.S. specifically, but are available for the broader North
America and Oceania (NAO) region. For example, many studies presented in this report rely on FLW data
developed by Gustavsson etal. (2011) for the Food and Agriculture Organization (FAO) of the UN, and this
data is broken down by global regions rather than individual countries. The NAO region, as defined by the
FAO, includes the U.S., Canada, Australia, and New Zealand.
While total FLW estimates for NAO are naturally larger than estimates for the U. S. alone, per person NAO
estimates are slightly lower than those for the United States. Compared to the other NAO countries, the U.S.
has a higher total and daily per person supply of calories, and a higher supply of meat per person, which has
an outsized impact on many of the resource inputs and impacts discussed in Chapter 4 (FAO, 2017; Roser and
Ritchie, 2013). Since the loss rates developed by Gustavsson et al. (2011) apply to all countries in NAO
(assuming an equal rate of FLW across countries), the food availability or supply drives FLW estimates (i.e.,
supply is the only variable differs among the countries). In terms of FLW, the per person values of FLW for the
U.S. are nearly the same as those of the NAO region (Chen et al., 2020; Guo et al., 2020).
Chapter 3. Characterization of U.S. Food Waste
25
-------
Grains
Fruits &
Vegetables
Meat & Fish &
Poultry Seafood
11
Total FLW by
Food Category
(relative weight)
Dairy &
Eggs
o
O °
O
Total FLW by
Supply Chain Stage
(relative weight)
NAO - FAO / Gustavsson et al. (2011)
U.S. - USDA / Buzby et al. (2014)
FIGURE 3-7. U.S. EDIBLE FLW BY RELATIVE WEIGHT, BY FOOD CATEGORY AND SUPPLY CHAIN STAGE
This figure compares food loss and waste (FLW) estimates, by food category and supply chain stage, from the Food and Agriculture
Organization (FAO) of the United Nations for the North America and Oceania (NAO) region to those from the United States Department of
Agriculture (USDA) for the United States only. The rows show FLW by supply chain stage, and the columns show FLW by food category.
FLW at the consumer stage is the greatest, followed closely by FLW at the primary production stage. By weight, fruits and vegetables are
the most lost and wasted food category.
Chapter 3. Characterization of U.S. Food Waste
26
-------
Several additional studies dive deeper and provide additional detail on the consumption stage. A study by Conrad
et al. (2018) built upon the USDA data by incorporating data on foods consumed (usually mixed dishes) from the
USDA What We Eat in America survey (a component of the National Health and Nutrition Survey). This study
examined the consumption supply chain stage (i.e., both at home and away from home) exclusively. The study
found fruits and vegetables (39 percent) and dairy products (17 percent), followed by meat (14 percent) and
grains (12 percent),20 to be the predominant components of food waste during the consumption stage, akin to the
findings of Buzby et al. (2014) shown above.
Two studies that directly measured household food waste - the OR DEQ's Oregon Wasted Food Study21
(McDermott et al., 2019) and the NRDC's assessment of three cites22 (Hoover and Moreno, 2017) - also support
the finding that fruits and vegetables (40 percent) are the largest components of consumption stage food waste,
followed by prepared foods and leftovers (23 to 28 percent). Dairy and egg, and meat and fish, however, were
found to be smaller contributors than in above studies (dairy and egg at 3 to 7 percent, and meat and fish at 6
percent), possibly due to some being counted in the prepared foods category. Results from the two studies are
presented in Figure 3-8. These data are from kitchen diaries kept by participating households. Trash sorts were
conducted by the researchers in the second study, and the sorts confirmed the findings from the kitchen diaries
(Hoover and Moreno, 2017).
Hoover & Moreno (2017)
Dairy & Egg
Fruits & Vegetables
Liquids,
Prepared Foods & Qj|s
Leftovers
Baked
Goods
"Snacks,
Sauces,
Meat & Condiments
McDermott et al. (2019)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
FIGURE 3-8. SHARE OF EDIBLE HOUSEHOLD FOOD WASTE, BY FOOD CATEGORY
This figure displays household food waste, by food category, recorded in kitchen diaries by participants in two U.S. studies.
In summary, across all stages of the food supply chain and during the retail and consumption stages, studies
agree that fruits and vegetables are wasted in the greatest amounts, typically followed by dairy and eggs. Other
animal products (i.e., meat and poultry) are wasted in smaller quantities than dairy and egg, according to most
studies. Seafood is the least wasted food category according to the data; however, this may be due to the
exclusion of many types of seafood losses during primary production (Love et al., 2015) and the relatively small
amount of seafood available at the retail level (Buzby et al., 2014).
20 The categories in Conrad et al. (2018) all consider the main ingredient in a prepared dish and thus include "mixed dishes."
For example, the fruits and vegetables category is defined as "fruits and vegetables, and mixed fruit and vegetable dishes".
21 Study covered of rural and urban areas of Oregon in 2017. 182 households completed 7-day kitchen diaries.
22 Study was conducted in Nashville, Denver, and New York City in 2016 and 2017. 613 households completed 7-day kitchen
diaries.
Chapter 3. Characterization of U.S. Food Waste
27
-------
3.7 Measurement Methodologies
The estimates of the environmental footprint of FLW presented in the subsequent chapters of this report rely upon
the data sets presented in this chapter to estimate FLW, including the amount and categories of food lost and
wasted and the supply chain stages at which the food was lost or wasted. Therefore, a general sense of the
methodological approaches used to create these data sets can assist in understanding the results of these
studies and in comparing results of multiple studies. FLW is typically estimated by comparing food availability to
either food utilization or energy requirements, though other methodologies are also available. This section
provides a brief overview of FLW measurement methodologies—those using a food balance approach
(comparing food availability and utilization), those using an energy balance approach (comparing food availability
and population's energy requirements), and those using mixed or other methods.
Food Availability & Utilization
Many FLW estimates rely upon a food balance approach, applying food loss and waste rates to estimates of food
availability. Food availability is the difference between the amount available in the commodity supply (i.e., the sum
of beginning stocks, production, and imports) and measured non-food uses (i.e., exports, farm and industrial uses
including seed and animal feed, and ending stocks) (IOM and NRC, 2015). Food loss and waste rates (often
called "loss rates" or "loss factors" regardless of supply chain stages covered) estimate the percent of available
food that is ultimately lost or wasted, by food category. The two primary sources of food availability data and food
loss and waste rates are:
1. U.S. Department of Agriculture (USDA) Economic Research Service (ERS)'s Loss-Adjusted Food
Availability (LAFA) data series (Buzby et al., 2014), and
2. Food and Agriculture Organization of the United Nations (FAO) Food Balance Sheets (Gustavsson et al.,
2011).
Many of the FLW estimates presented in this report rely on one of these data sources for food availability data or
food loss and waste rates, or both. For example, many of the studies focused exclusively on the United States
utilize USDA data, while studies seeking to compare regions or countries most often rely upon the FAO data.
More details on studies' reliance on USDA or FAO data can be found in Appendix B.
Both the USDA and FDA data sources provide annual total and per person food availability data for a wide variety
of food categories (USDA data breaks food into 200 categories, while FAO uses 100 categories). The FAO
estimates cover a larger geographic area (the NAO region) and scope (FLW during all four stages of cradle-to-
consumer supply chain), than the USDA estimates (U.S. FLW at retail and consumption stages only) (Buzby et
al., 2014; Gustavsson et al., 2011). Both sources also provide loss rates - the FAO loss estimates developed by
Gustavsson et al. (2011), or the USDA's Loss Adjusted Food Availability (LAFA) data series (Buzby et al., 2014) -
which can be applied to food availability data to calculate FLW. The loss rates from the two sources differ, and
this (along with differences in availability estimates) impacts their estimates of FLW.
As shown in Figure 3-9, FAO (Gustavsson et al., 2011) and USDA (Buzby et al., 2014) loss rates differ
substantially for some food categories. For example, while FAO assumes 42 percent of fruits and vegetables and
11 percent of meat and poultry are wasted during the consumption stage, USDA estimates 21 percent of each
category are wasted.
After applying loss rates, the USDA per person estimates of FLW from the retail and consumption stages, and for
consumption stage alone (229 kg and 132 kg, respectively) are higher than those of FAO (140 kg and 118 kg,
respectively), which can impact the estimates of inputs and environmental impacts presented in Chapter 4 (Buzby
et al., 2014; Gustavsson et al., 2011). This can be viewed in the rightmost column of Figure 3-7 in the previous
section (Section 3.6).
Chapter 3. Characterization of U.S. Food Waste
28
-------
®T\ i?\ © \ <®>
'rimary J Distribution & Processing J t .. J
Auction J °nS"m"
Fish & |
Seafood
Meat & ¦ 4%
Poultry
Dairy & |g 4o/o
Eggs
Grains
12%
Fruit & |
Vegetables
| 4%
| 1%
| 1%
1 2%
¦ 4%
6%
14%
| 1%
I 2%
I
4%
4%
1%
11%
12%
31%
Y/////S//S//A
42%
|FAO/Gustavssonetal. (2011)
USDA/Buzbyetal. (2014)
F = fresh; M = milled; P = processed
FIGURE 3-9. COMPARING FAO AND USDA LOSS RATES
This figure compares the loss rates (i.e., the percent of available food estimated to be lost or wasted) by food category
and food supply chain stage from the Food and Agriculture Organization (FAO) of the United Nations and the United States
Department of Agriculture (USDA). The USDA only estimates retail and consumption stage losses. Loss rates at the
consumption stage, from both sources, include cooking losses.
Researchers must decide which food parts are edible, how they will quantify inedible parts, and to which supply
chain stage they will attribute the waste (i.e., removal) of inedible parts when developing FLW estimates. In both
the FAO and USDA data sets, researchers estimated how much of a given food group is typically edible and
applied this conversion factor to the amount of food available to consumers from that food group. For example, to
create the FAO estimates of edible FLW, Gustavsson et al. (2011) applied a conversion factor of 0.77 to all fruit
and vegetable FLW to calculate and remove the portion that was inedible from the loss estimate. This results in
23 percent of the weight of fruits and vegetables being considered inedible (Gustavsson et al., 2013). Gustavsson
et al. (2011) did not have conversion factors for meat and dairy, so the weight of bones is included in the edible
values. By comparison, USDA uses data from the National Nutrient Database for Standard Reference that details
the inedible portion of thousands of foods (USDA, 2018). For example, apples, broiler chickens, and broccoli are
considered to be 10, 20, and 39 percent inedible by weight, respectively. Within the USDA data, inedible shares
are removed at different stages within the food supply chain. For example, for meat and poultry the retail weights
reflect the edible weight, but for fresh fruits and vegetables the retail weight includes the inedible portions and
those are removed prior to the consumer weight (Buzby et al., 2014).
Chapter 3. Characterization of U.S. Food Waste 29
-------
Food Availability & Energy Requirements
An alternative approach to estimating FLW is to subtract a populations' energy requirements (i.e., a surrogate for
food consumption) from an estimate of food availability. Based on human metabolism models, researchers
quantify the amount of energy, in the form of calories, that is needed to maintain a population's current physical
activity levels and body weights.23 Thus, the estimates must include assumptions on activity levels and
metabolism and convert food tonnages to calories and nutrients. This approach, used by four studies presented in
this chapter (Lopez Barrera and Hertel, 2020; van den Bos Verma et al., 2020; Hi
-------
CHAPTER 4.
Environmental Footprint of
U.S. Food Loss and Waste
Over the past decade, more than a dozen studies have assessed the
environmental footprint of producing, storing, processing, packaging,
distributing, and marketing food that is ultimately lost or wasted.
Studies typically examine the use of resources (land, water, and
energy) and other inputs (pesticides and fertilizers) as a proxy for the
environmental impacts their use may cause (e.g., deforestation,
water scarcity, or decreased water quality). Typically, GHG emissions
are the only output of FLW directly estimated in the literature. Many
of the other farm to kitchen environmental impacts of FLW discussed
in Chapter 2, such as biodiversity loss, soil degradation, and GHG air
emissions, are not quantified in the literature. While the
environmental impacts may be considered a trade-off to an abundant
supply of food, the impacts associated with food loss and waste are
in many cases completely unnecessary and could be avoided.
Throughout the chapter, summary tables are provided for each input
and environmental impact, with an analysis of all available estimates
and a recommended value for policymakers to use in communicating
the environmental footprint of FLW and decision-making among
competing priorities (including FLW). Details about supply chain
stages and food categories that contribute the most to each input or
impact are also provided to assist policymakers in designing targeted
FLW reduction strategies.
4.1 Methodologies
Quantifying the farm to kitchen environmental footprint of FLW
requires data on the amount and categories of food that are lost or
wasted at each stage of the food supply chain (see Chapter 3) plus
the environmental footprint of FLW at that stage and all previous
stages (see Chapter 2). The impacts are cumulative; for example,
food lost during primary production embodies the resources used to
grow the food, whereas food wasted during the consumption stage
embodies the resources used from the primary production stage to
the point the food reached the consumer. A simplified depiction is
presented in Figure 4-1.
KEY FINDINGS
¦ Each year, U.S. FLW
embodies:
o Agricultural land:
560,000 km2
(140 million acres)
o Blue water:
22 trillion L
(5.9 trillion gallons)
o Fertilizer:
6,350 million kg
(14 billion pounds)
o Energy:
2,400 million GJ
(664 billion kWh)
o GHG emissions:
170 million MTCC>2e
GHG (excluding
landfills)
¦ Inputs (e.g., land, water,
fertilizer, or energy) are
typically examined as a
proxy for the
environmental impacts
their use may cause
(e.g., deforestation, water
scarcity, decreased water
quality, or climate
change)
¦ Animal products have an
outsized contribution to
the environmental
footprint of U.S. FLW,
representing the greatest
use of resources (land,
water, fertilizer, energy)
and GHG emissions
among categories of
FLW, but a relatively
small share of FLW.
Fruits and vegetables are
also leading contributors.
Chapter 4. Environmental Footprint of U.S. Food Waste
31
-------
§4
Primary
Production
Distribution a
& Processing A
V
Consumption
Amount & Category
of FLW
Amount & Category
of FLW
Amount & Category
of FLW
Amount & Category
of FLW
T
Environmental
Impacts
Environmental
Impacts
Environmental
Impacts
Environmental
Impacts
FIGURE 4-1. ESTIMATING THE CRADLE-TO-CONSUMER ENVIRONMENTAL FOOTPRINT OF U.S. FLW
Scope
The studies presented in this chapter rely on estimates and characterizations of FLW presented in Chapter 3. Key
variables among the FLW estimates underlying the studies include:
Edibility (i.e., whether the estimates include inedible food parts, such as bones and shells)
Categories of food included (i.e., whether the estimate includes animal products and the feed
needed to produce them)
Supply chain stages (i.e., whether FLW from each supply chain stage was included)
Geographic area (i.e., whether the estimate is U.S.-specific or for a broader region)
These factors can dramatically influence the magnitude of environmental estimates. For more detailed discussion
of these variables, see Sections 3.4 to 3.6 including the text box about the NAO region.
Chapter 4. Environmental Footprint of U.S. Food Waste 32
-------
Approaches
Two key methods are used to estimate the inputs and environmental impacts associated with FLW- life cycle
assessment (LCA) and environmentally extended economic input-output models (EEIO). Life cycle assessment
(LCA) takes a detailed look at the inputs and environmental impacts attributable to a product within the defined
scope of the analysis (e.g., cradle-to-consumer). Within an LCA, life cycle inventory data are used to quantify the
resource inputs (such as energy, water, and land use) and resulting outputs (such as emissions and wastes) for
each stage of the product's life (Muth et al., 2019). Depending on the purpose of the study, LCAs can use national
average data, data from multiple primary sources, or from a single source. When using LCA for estimating the
environmental impacts of diets and food systems researchers might use an average across multiple LCAs to
represent a single product or results from a single LCA. When LCAs of single products or commodities are
compiled to represent activity at a broader level (e.g., nationally), some resolution may be lost (e.g., assuming
that the LCA of one vegetable grown in Idaho represents all vegetables in that category grown in the U.S.).
In contrast, EEIO models map a network of relationships based on monetary transactions between industry
sectors and their resulting products at an economy-wide scale. EEIO models incorporate life cycle inventory data
into the input-output framework and enable the calculation of embedded direct and indirect inputs and
environmental impacts of products (e.g., food). For example, assume that Industry A creates 100 widgets and, in
the course of doing so, generates 100 kg of a concerning pollutant. If Industry B purchases 25 percent of the
widgets, then 25 kg of the pollutant is associated with the demand created by Industry B. These 25 kg are
"embedded" in the output produced by Industry B. When Industry C uses the product produced by Industry B, the
25 kg are "passed on" to the products of Industry C. In this way, impacts that occur throughout the U.S. are
allocated to products (U.S. EPA, 2020e).
EPA's U.S. Environmentally-Extended Input-Output (USEEIO) Model (Yang et al., 2017), used in the study by
Read et al. (2020). EPA's U.S. Environmentally-Extended Input-Output (USEEIO) Model (Yang et al., 2017), used
in the study by Read et al. (2020) presented in this chapter, calculates environmental impacts and resource use at
a national scale using publicly-available data. This capability comes with limitations. The most important is that
the data are somewhat aggregated and coarse. Thus, the level of resolution is currently limited to national
averages for a sector or product classification. However, the 26 aggregated agriculture and food manufacturing
industries in the USEEIO provide a mechanism for examining the U.S. food system and the inputs and
environmental impacts associated with FLW. EEIO models offer the advantage of capturing indirect impacts, such
as those from producing equipment (which are often not included in process-level LCAs), as food moves along
the supply chain. Thus, they may be expected to generate higher estimates than LCA studies (Heller and
Keoleian, 2015). However, when one supply chain stage (e.g., primary production) is responsible for the majority
of an impact (e.g., use of land, fertilizers, or pesticides), process-level models may be more useful and
representative than sector-based models.
Other Considerations
A key limitation of all the studies presented in this chapter is the lack of accounting for differences in inputs and
environmental impacts of imported foods (see Section 2.4). Most studies discussed in this chapter assume the
U.S. food supply was produced entirely domestically and that U.S. average environmental factors apply. A couple
- Chen et al. (2020) and Skaf et al. (2021) applied the same international lifecycle assessment and production
factors for all the countries in the study, including the U.S. This may over- or underestimate the environmental
inputs and impacts of the foods, depending upon where they originate. There are two excepted estimates, one is
ReFED's (2021a) greenhouse gas impacts value and the other is the energy value associated with FLW from
companion studies by Pagani et al. (2020), and Vittuari et al. (2020). ReFED's (2021a) analysis, conducted by
Quantis and utilizing their internal life cycle inventory database, accounted for imports by matching the top
producing countries for each food item with the available country-specific production data. Pagani et al. (2020)
and Vittuari et al. (2020) excludes imports from the analysis due to the lack of reliable data on the energy
embodied in foreign production and international transport, thus underestimating inputs and environmental
impacts, since imports account for up to one-fifth of the U.S. food supply (USDA, 2019f).
The use of national average environmental factors may also over- or underestimate the environmental inputs and
impacts of specific domestically produced foods, in cases where inputs and impacts vary widely within food
groups or based upon production method, geographic location, seasonality, type and amount of processing
conducted or packaging used, whether cold storage is required, or type and distance of transportation (see
Section 2.5). When targeting FLW initiatives to maximize environmental benefits, it may be useful to consider
these factors.
Chapter 4. Environmental Footprint of U.S. Food Waste
33
-------
4.2 Agricultural Land Use
Land is a limited resource with many competing uses. While the U.S. has more suitable land for growing crops
than most other countries, the amount of land used for growing food and animal feed in the U.S. has slightly
declined since 1982, while developed land has increased and timberland has remained constant (U.S. EPA,
2020c). However, land use changes may occur in countries producing food for import into the U.S. and can
significantly impact water quality, release GHG emissions (as carbon is held in healthy soils and trees), and result
in deforestation and the loss of biodiversity and ecosystem services. In addition, as the global population rises
and diets shift due to rising incomes, more land will be needed to produce food. Land "wasted" by producing food
that is ultimately lost or wasted could instead reduce the conversion of more land to cropland.
Seven studies estimated the amount of land required to produce food that was ultimately lost or wasted in the
U.S. Table 4-1 shows the values from these recent studies. Six of the studies (Chen et al., 2020; Conrad et al.,
2018; Birney et al., 2017; Toth and Dou, 2016; Kummu et al., 2012) used a similar approach to calculate
agricultural land use associated with FLW. These studies evaluated current diets and amounts of calories
consumed or lost in the food system and multiplied the food categories by land use characterization factors to
calculate the amount of land used in production of those foods. Read et al. (2020) instead used EEIO models, as
described in Section 4.1 to quantify the inputs and environmental impacts of FLW using consumer expenditures.
The authors first modified the EEIO model by Miller and Blair (2009) to represent the U.S. food supply chain and
then used EPA's USEEIO model to estimate the embodied environmental inputs in FLW.
The estimate from Read et al. (2020) addressed the broadest scope of any study, including FLW from all stages
of the food supply chain and including land used for animal feed and livestock grazing. As such, it is not surprising
that the authors' estimate (561,000 km2 per year) is larger than those from studies that only included FLW from
part of the food system (Conrad et al., 2018; Birney et al., 2017; Toth and Dou, 2016) or did not include all land
use related to livestock production (Chen et al., 2020; Kummu et al., 2012). Kummu et al. (2012), for example,
included only crop products intended for direct human consumption, meaning animal products and the feed crops
needed to produce them were both excluded throughout the study. Chen et al. (2020) excluded only pasture land.
While pasture lands typically have no potential to become cropland, some studies include them due to potential
negative effects on carbon storage and biodiversity (Bajzelj et al., 2014). Studies also differed in the scope of
FLW included, and studies that estimated larger scopes of FLW generally estimate greater inputs and
environmental impacts. Despite differences in modeling approaches and methods for estimating food waste,
estimates from all the reviewed studies are in relative agreement once these differences in scope are considered.
While all agricultural land use occurs during primary production.
Note that a key limitation of all the estimates is the lack of accounting for differences in the environmental impact
of imported foods (see Section ome of the land use quantified here may relate to deforestation or loss of
biodiversity in the producing country. Also, assuming food that was ultimately lost or wasted was produced on the
amount of land that would be used in the U.S. to produce it likely underestimates agricultural land use, as the U.S.
has some of the world's most productive agricultural lands in the world (Conrad et al., 2018).
In summary, the agricultural land use estimate from Read et al. (2020) is the most comprehensive available, and
thus may be the most useful to policymakers. Read estimates that 560,000 km2 (140 million acres), or 1,800 m2
(19,000 sq ft) per person, are used annually to produce food that is ultimately lost or wasted. This is equivalent to
approximately 16 percent of agricultural land in the U.S. (including harvested and unharvested cropland,
rangeland, and pastureland).
Chapter 4. Environmental Footprint of U.S. Food Waste
34
-------
TABLE 4-1. ANNUAL AGRICULTURAL LAND USE ASSOCIATED WITH U.S. FLW
Land Use
Source
Scope of FLW
Total
(km2)
Per Person
(m2/person)
Scope of
Land Use
Land Use
Factors
Read et al. (2020)
560,000*
1,800 m Ea us-
Used EPA's USEEIO model used to estimate land use associated with FLW.
Kummu et al. (2012)
178,000*
498
Intl
Excludes loss and waste of animal products. Calculated the national cropland
yield by commodity.
Toth and Dou (2016)
260,000
830*
U.S.
Excludes land for orchard fruit and nuts and perennial forage crops. Based on
USDA National Census Data.
Birney et al. (2017)
325,500*
1,051
U.S.
Excludes land use to produce dairy. Includes only harvested cropland. Used
FAOSTAT 2010 yield data. Supplemented with land requirements for poultry,
eggs, pork, beef, and lamb from studies of New York State and North Carolina.
Chen et al. (2020)
HI
118,000*
378
Intl
Matched recently available global average characterization factors per food group
(i.e., cropland use [m2/g]) to product resolution.
Conrad et al. (2018)
in
120,000
390* U.S.
Used the U.S. Foodprint Model which models the U.S. as a closed food system.
Skaf etal. (2021)
in
198,000
606.6
Intl
Used LCA data from Ecolnvent and applied the ReCiPe midpoint method for
agricultural land occupation.
In addition to cropland for direct human consumption, indicates inclusion of cropland for animal feed and indicates inclusion of pasture and rangeland.
= Indicates estimate for the NAO Region rather than for the U.S. specifically.
• = Calculated values for total and per person used the same population factor as the study's data year.
f= personal communication with the author.
Chapter 4. Environmental Footprint of U.S. Food Waste
35
-------
Agricultural Land Use, By Food Category
While Read et al. (2020) provided the most comprehensive estimate of land use associated with FLW, other
studies with smaller scopes provide greater insights into the food categories most responsible for the agricultural
land use associated with FLW. As shown in Chapter 3, many studies rely on the detailed data on U.S. FLW by
food category from USDA, which is available for the retail and consumer stages only.
Toth and Dou (2016), Birney et al. (2017), and Conrad et al. (2018) all provide estimates of agricultural land use,
by category of FLW. This information may be useful for policymakers desiring to curb agricultural land use and its
potential environmental impacts through FLW initiatives. For estimates of land use from producing animal
products, each of the studies included land required to produce enough animal feed to support production of the
animal-based foods.
As shown in Figure 4-2, the three studies found the vast majority of land use associated with FLW was
attributable to animal products (including the land used to grow hay, feed grains, and oilseeds). This
demonstrates that the loss and waste of animal products has an outsized effect on land use. While they represent
only 30 percent of FLW along the supply chain (FAO, 2011), they account for roughly two-thirds of agricultural
land use associated with FLW (Conrad et al., 2018; Birney et al., 2017; Toth and Dou, 2016). This also explains
the large difference between agricultural land use estimates of Read et al. (2020) and Kummu et al. (2012) in the
previous section, as Kummu et al.'s study excludes land used to produce animal products (including animal feed)
and is roughly one-third of Read et al.'s estimate, which includes animal products and animal feed.
Animal Feed
Direct Human
Consumption
Toth & Dou (2016)
I I
Sweeteners—I
Fruit —
I
I— Grains
I— Vegetables
Birney et al. (2017)
Conrad et al. (2018)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Agricultural Land Use
FIGURE 4-2. AGRICULTURAL LAND USE ASSOCIATED WITH U.S. FLW, BY FOOD CATEGORY
Chapter 4. Environmental Footprint of U.S. Food Waste ^h 36
-------
4.3 Water Use
Freshwater is a vital, natural resource used every day by plants, animals, people, and industries. The extent of
water resources (their amount and distribution) and their condition (physical, chemical, and biological attributes)
are critical to ecosystems, human uses, and the overall function and sustainability of the hydrologic cycle. When
food is lost or wasted, so too is the water used to grow and produce it. Nine studies estimated the amount of
water wasted from producing uneaten food in the U.S. Most studies utilize a bottom-up approach, where
researchers assess how much of each food category was wasted and then apply factors approximating how
much water is used to produce each unit of food in that category. Table 4-2 and Table 4-3 summarize the results.
All the studies presented in this chapter measure blue water use during primary production (e.g., for irrigation and
livestock watering). Some measure only irrigation (Conrad et al., 2018; Toth and Dou, 2016; Kummu et al.,
2012),26 while others include livestock watering (Chen et al., 2020; Birney et al., 2017). Only Read et al. (2020)
also accounts for uses of blue water during other supply chain stages, such as during food processing and food
preparation;27 thus, the other studies underestimate the cradle-to-consumer blue water footprint of FLW. As
context, Canning et al. (2020) estimates that the primary production stage represents 65 percent of the water use
of the cradle-to-consumer food system, with food distribution and processing an additional 3 percent and
consumption accounting for another 20 percent.
Looking at water use throughout the entire cradle-to-consumer food system, Read et al. (2020) estimated 22
trillion liters of water use annually from FLW along the entire supply chain. Estimates from the other studies,
which measured only blue water use during primary production, ranged from 11 to 53 trillion L, consistent with
one another and lower than Read et al.'s (2020) estimate as would be expected. The highest estimate, from
ReFED (2021a), stands out because of its similarities to Birney et al. (2017) yet higher end result. Both assessed
similar amounts of FLW (73 and 71 million metric tons, respectively) and relied on blue water use factors for food
production from Mekonnen and Hoestra (2011, 2012), but ReFED also included water use for additional supply
chain stages, like food processing and manufacturing.
In addition to blue water use, agricultural production also utilizes green water flows (i.e., rainwater that is soaked
up, staying on vegetation or in the soil) (Mekonnen and Hoekstra, 2011). Thus, the values for blue water use
presented in the studies understate the full water footprint associated with FLW. Two studies (Mekonnen and
Fulton, 2018; Birney et al., 2017) estimated green water use associated with FLW during the retail and
consumption stages, finding that green water represents 88 percent of total water use (i.e., use of blue water plus
use of green water) during primary production.
In summary, the blue water use estimate from Read et al. (2020) is the most comprehensive available. The
authors estimate that all FLW is responsible for 22 trillion L (5.9 trillion gallons) of blue water use, or 71,000 L
(19,000 gallons) per person, annually. This is equivalent to the annual blue water use of more than 50 million
American family homes (U.S. EPA, 2018).
26 Kummu et al. (2012) excluded irrigation of animal feed.
27 Read et al. (2020) uses the USEEIO model which captures both direct and indirect resource use along the supply chain.
Chapter 4. Environmental Footprint of U.S. Food Waste
37
-------
TABLE 4-2. ANNUAL BLUE WATER USE ASSOCIATED WITH U.S. FLW
Water Use - Blue Water
Scope of FLW
Total
(trillion L)
Per
Person
(L/year)
Scope of
Water Use
Water
Use
Factors
ReFED (2021a)
>«>
163,000
i%i
U.S.
Used blue water factors from the Water Footprint Network
which are based on water factors for production from
Mekonnen & Hoekstra (2011 & 2012).
BC33S
Wi
Iktfl
Read et al. (2020)
EDXiiXiil
71,000
U.S.
Used EPA's USEEIO model to estimate blue water
consumed to produce wasted food. Included water use
along cradle-to-consumer food supply chain.
Kummu et al. (2012)
>$>
42,000
NAO
Used data aggregated from NAO region to calculate the
irrigation water used to produce vegetative food waste.
Toth and Dou (2016)
Used USDA-NASS irrigation survey data.
U.S.
Birney et al. (2017)
54,000
U.S.
Used Mekonnen and Hoekstra (2011) life cycle analysis of
the blue and green water requirements for food.
Mekonnen and
Fulton (2018)
33,277
U.S.
Used Mekonnen and Hoekstra (2011) for crop products;
Mekonnen and Hoekstra (2012) for animal products, life
cycle analysis of the water requirements for food products.
Chen et al. (2020)
in
54,930
Iritl
Matched available global average characterization factors
per food group (i.e., water use [l/g]) to product resolution.
Conrad et al. (2018)
Hi
51,000'
U.S.
Used USDA-NASS farm and ranch irrigation survey data
and applied those rates to the estimates of cropland
associated with FLW.
Skaf et al. (2021)
HI
33,950
Intl
Used LCA data from Ecolnvent and applied the ReCiPe
midpoint method for water depletion
= Indicates estimate for the NAO Region rather than specific to the U.S.
* = Calculated values for total and per person used the same population factor as the study's data year.
= personal communication with the author.
Chapter 4. Environmental Footprint of U.S. Food Waste
38
-------
TABLE 4-3. ANNUAL GREEN WATER USE ASSOCIATED WITH U.S. FLW
Source
Scope of FLW
Water Use - Green Water
Total
(trillion L)
Per Person
(L/year)
Scope of
Water Use
Water Use
Factors
Birney et al. (2017)
397,000
U.S.
Used Mekonnen and Hoekstra (2011) life cycle analysis
of the blue and green water requirements for food items.
Mekonnen and Fulton
(2018)
247,400
U.S.
Used Mekonnen and Hoekstra (2011) for crop products;
Mekonnen and Hoekstra (2012) for animal products, life
cycle analysis of the blue and green water requirements
for food products
: calculated value used the same population factor as the study's data year.
Water Use, By Food Category
Three studies (Conrad et al., 2018; Mekonnen and Fulton, 2018; Birney et al., 2017) provide details by food
category on blue water used to grow food that was ultimately wasted during consumption stage (Conrad et al.,
2018) or the retail and consumption stages (Mekonnen and Fulton, 2018; Birney et al., 2017). Figure 4-3 shows
the proportion of blue water used for each FLW food category.
All three studies showed animal products accounting for approximately a third to more than half of the water use
associated with FLW. Fruits and vegetables also accounted for substantial water use, from roughly one-fifth to
more than half of all water use associated with FLW. ReFED (2021a) also calculated water use by food category,
but grouped foods by their retail form like frozen, prepared foods, and dry goods. Even with different groupings,
ReFED found that fresh meat and seafood accounted for 30 percent of the water associated with FLW, whereas
produce comprised 7 percent. A couple other studies (Read et al., 2020; Toth and Dou, 2016) similarly identified
the waste of animal products and fruits and vegetables as having substantial contributions to the blue water
footprint of food waste. Note that these analyses (other than Read et al. (2020) and ReFED (2021a)) do not
include water use beyond primary production, such as during food processing or food preparation.
Direct Human Consumption
Mekonnen & Fulton
(2018)
Fruits
Vegetables Grains
Birney et al. (2017)
Conrad et al. (2018)
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Water Use
FIGURE 4-3. BLUE WATER USE ASSOCIATED WITH U.S. FLW, BY FOOD CATEGORY
Chapter 4. Environmental Footprint of U.S. Food Waste ^¦
39
-------
4.4 Pesticide and Fertilizer Application
Application of pesticides and fertilizers help increase the productivity of croplands by reducing loss from insects
and plant disease and ensuring that crops have the essential nutrients they need to grow; however, they can also
have unintended consequences when they migrate from croplands to water bodies and surrounding areas. This
section examines the use of pesticides and fertilizers during the primary production of food that is ultimately lost
or wasted.
Pesticide Application
Only one study examined pesticide use associated with food that was ultimately lost or wasted. Conrad et al.
(2018) estimated pesticide application associated with consumption stage FLWto be 354 million kg (778 million
pounds) of pesticides annually, equivalent to 1 kg (2.5 pounds) per person per year, as shown in Table 4-4. No
studies were available examining pesticide use associated with FLW during the other supply chain stages. For
context, the consumption stage accounts for approximately half of U.S. FLW (Pagani et al., 2020; Vittuari et al.,
2020; CEC, 2017).
TABLE 4-4. ANNUAL PESTICIDE APPLICATION ASSOCIATED WITH U.S. FLW
Source
Scope of FLW
Pesticide Application
Total
(million kg)
Per Person
(kg/person)
Scope of
Pesticide
Application
Application
Rates
Conrad et al. (2018)
HI
350
1* ^ U.S.
Used USDA National Agricultural Statistics Service agricultural
survey data.
In addition to pesticides applied to cropland grown for direct human consumption, indicates inclusion of pesticides applied to cropland
for animal feed. • = calculated value used the same population factor as the study's data year.
Pesticide Application, By Food Category
Conrad et al. (2018) provided additional information on the pesticide application from each food category of
consumption stage FLW. As shown in figure 4-4, fruits and vegetables account for more than one half of pesticide
application among all food categories of FLW. An additional one quarter of wasted pesticides were applied to feed
grains, oilseeds, and hay to support animal production. All other FLW accounted for the remaining 13 percent of
pesticide application.
Animal Feed Direct Human Consumption
I II 1
Sweeteners Fruits Vegetables Grains
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Pesticide Application
FIGURE 4-4. PESTICIDE APPLICATION ASSOCIATED WITH U.S. FLW, BY FOOD CATEGORY
Chapter 4. Environmental Footprint of U.S. Food Waste
40
-------
Fertilizer Application
There are three major types of commercial fertilizer used in the U.S. - nitrogen, phosphate, and potassium (or
potash). Nutrient runoff from nitrogen (N) and phosphorous (P) fertilizers can lead to eutrophication and algal
growth in water bodies, and N fertilizer can also stimulate the release of nitrous oxide, a GHG, from soils
(Davidson, 2009). When used at recommended application rates, there are few to no adverse effects from
potassium (K) fertilizer. Where possible, the data on fertilizer application in this report is broken down into these
components since the environmental impacts vary. Six studies assessed the amount of fertilizers used to produce
food that was ultimately lost or wasted. To calculate estimates, the authors applied fertilizer application rates to
estimates of the land used to produce the FLW. Table 4-55 summarizes the results.
Toth and Dou (2016) estimated 6.35 billion kg of fertilizer were used to grow food that ultimately was wasted. The
authors' estimates were based on USDA-NASS survey data on crop-type specific fertilizer application rates and
the percentage of acres fertilized for food and feed production. Examining the application of specific nutrients,
Toth and Dou (2016) estimated 2.7 billion kg of nitrogen application and 1.5 billion kg of phosphorus application
were associated with the production of FLW annually.
While the estimate of Toth and Dou (2016) is not comprehensive, since it excludes FLW during primary
production, it is the most complete estimate available in the literature. Other studies presented in Table 4-5
exclude fertilizer use on animal feed crops (Birney et al., 2017; Kummu et al., 2012) or evaluate a more limited
scope of FLW (Conrad et al., 2018; Birney et al., 2017), which would likely lead to underestimates of FLW.
Estimates from the other studies are lower than that of Toth and Dou (2016), each at a scale roughly consistent
with their smaller scope.
In addition, several of the studies relied on international rather than U.S.-specific fertilizer application rates (Chen
et al., 2020; Birney et al., 2017; Kummu et al., 2012) which may affect the precision of the studies' estimates.
According to FAO data, between 2002 and 2017, U.S. application rates (i.e., amount per unit of land) of
phosphorous were consistently lower (by 12 to 27 percent) than the global average, while U.S. application rates of
nitrogen were similar (3 percent higher) to the global average (Our World in Data, 2021). Thus, studies using
international factors may over-estimate phosphorous application.
Another study, Read et al. (2020) includes FLW along the supply chain; however, it is not directly comparable to
that of Toth and Dou (2016) or other studies. Rather than producing an estimate of fertilizer use, Read et al.
(2020) estimates the "eutrophication potential" of the nitrogen fertilizer use associated with FLW. The authors use
the life cycle impact indicator "eutrophication potential" by applying a fate and transport model to the amount of
nitrogen fertilizer applied details not provided in the study) to cropland used to grow food that was ultimately
wasted. The authors calculated eutrophication potential associated with FLW of 1.7 kg of nitrogen equivalent per
person (i.e., that amount of nutrient reached a water body thereby impacting the water quality). While this metric
may ultimately be more useful in estimating environmental impacts than simply quantifying inputs, details were
sparse and a scale upon which to judge the value presented was not provided. Similarly, Skaf et al. (2021)
calculated a freshwater eutrophication value of 0.51 kg phosphorus equivalent per person associated with
consumer food waste.
In summary, Toth and Dou provide the most comprehensive estimate of the application of fertilizer associated
with FLW-6.35 billion kg of fertilizer (14 billion pounds) or 20.2 kg (44.5 pounds) per person, annually. This is
equivalent to the average amount of fertilizer used on 100 million acres (U.S. EPA, 2019a). The fertilizer
estimates can be broken down into elements, showing an estimated 8.5 kg per person of nitrogen application and
4.7 kg per person of phosphorus application each year.
Chapter 4. Environmental Footprint of U.S. Food Waste ^h 41
-------
TABLE 4-5. ANNUAL FERTILIZER APPLICATION ASSOCIATED WITH U.S. FLW
Fertilizer Application
Source
Scope of FLW
Nitrogen (N)
Phosphorus (P2O5)
Total Fertilizer
Sum of N, P2O5 & K2O
Total
(million kg)
Per Person
(kg/person)
Total
(million kg)
Per Person
(kg/person)
Total
(million kg)
Per Person
(kg/person)
Scope of
Fertilizer
Application
Application
Rates
Kummu et al. (2012)
3,300*
9.3
Intl
Based on national-level cropland area divided by national-level
fertilizer use.
Toth and Dou (2016)
2,670*
8.5*
1,460*
6,350
20.2*
U.S.
Based on USDA-NASS census data of average annual fertilizer
application rates and percent of acres fertilized for food and animal
feed by crop type.
Birney et al. (2017)
5,300*
17
Intl
Used fertilizer consumption data from the International Fertilizer
Industry Association and land use data from FAOSTAT.
Chen et al. (2020)
in
2.7
150*
4.7
0.5 - ~ WL lntl
Fertilizer amounts determined by matching recently available
global average characterization factors per food group (i.e., N and
P2O5 application [g/g]) to product resolution.
Conrad et al. (2018)
HI
2.6*
2.2*
2,500
8*
U.S.
Used USDA-NASS Agricultural Survey data.
In addition to fertilizer application on cropland for direct human consumption, indicates inclusion of cropland for animal feed. Q = Indicates estimate for the NAO Region rather than for the U.S.
specifically. • = Calculated values for total and per person used the same population factor as the study's data year. For Toth and Dou (2016) values for total nitrogen and total phosphorus, the
percentages reported by the authors total fertilizer application were applied to the authors' total for fertilizer associated with FLW.
Chapter 4. Environmental Footprint of U.S. Food Waste
42
-------
Animal Feed Direct Human Consumption
ii
i
Total (NPK)
Sweeteners
Fruits Vegetables Grains
Nitrogen
¦
Phosphorus
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Fertilizer Application
FIGURE 4-5. FERTILIZER APPLICATION ASSOCIATED WITH U.S. FLW, BY FOOD CATEGORY
Data Source: Conrad et al. (2018)
Fertilizer Application, By Food Category
In addition to estimating land use and pesticide application, Conrad et al. (2018) provides estimates of fertilizer
application associated with consumption stage FLW by food category. As shown in Figure 4-5, the authors
estimate that the largest share of nitrogen fertilizer application (over 40 percent) was from the production of feed
grains, oilseeds, and hay grown to support animal products that were ultimately wasted. Conrad et al. (2018) also
finds that fruits and vegetables comprise a substantial share (approximately 30 percent) of nitrogen fertilizer
application. Examining phosphorous application, Conrad et al. (2018) finds similar results. The largest share of
phosphorous application (almost 60 percent) is from feed grains, oilseeds, and hay grown to support animal
products that were ultimately wasted, and the next largest share is attributable to fruits and vegetables (almost 25
percent). A study by Wood et al. (2019) confirms the prominence of animal products as a contributor to
phosphorous fertilizer application and to ammonia air emissions (from nitrogen fertilizer application) associated
with FLW.
This is consistent with the finding in Section 4.2 that the largest share of land use associated with FLW is
attributable to animal products (and the feed grown to support them) and that these products comprise an
outsized portion of inputs compared to their share of FLW, which is approximately 30 percent (FAO, 2011). In
contrast, fruits and vegetables represent a slightly higher percentage of FLW (more than 40 percent) (FAO, 2011).
The work of Chen et al. (2020), while potentially less precise due to its use of international fertilizer application
rates, confirms this finding. They found that, for the NAO region, animal products were responsible for 26 percent
and 31 percent of nitrogen and phosphorus use, followed by cereals (24 percent of N and 21 percent P) and fruits
and vegetables (20 percent of N and 19 percent of P).28
28 Chen et al. (2020) estimates are not included in Figure 4-3 since the study broke down FLW into different categories than
Conrad et al. (2018), thus making comparison of the two studies' results difficult. Conrad et al. (2018) was selected over Chen
et al. (2020) due to its use of U.S. application factors.
Chapter 4. Environmental Footprint of U.S. Food Waste
43
-------
4.5 Energy Use
To describe the energy inputs that are used along the food supply chain to produce food for consumers,
researchers often quantify the "embodied energy" of food types. Embodied energy is the cumulative amount of
energy that was used to produce a food product through a given stage in the food system (e.g., from cradle-to-
consumer). The further along the food supply chain, the higher the level of embodied energy because it is a
summation of all the earlier inputs. Unlike the use of land or chemicals, which predominately occur on-farm,
energy use occurs all along the supply chain, so the embodied energy of a food product is considerably higher at
the consumption stage than during primary production.
Table 4-6 presents estimates of the embodied energy of U.S. FLW. All studies included energy use along the
entire food supply chain. Unlike estimates for inputs such as water, these estimates are created using a top-down
approach, starting with total U.S. energy consumption, typically from the U.S. Energy Information Administration
(EIA), then distributing it to sectors (e.g., distribution and processing, or retail) and their functions (e.g.,
transportation or refrigeration), then associating the appropriate portion to FLW.
The analysis by Cuellar and Webber (2010) brought to light the large amounts of energy used by the food system
and the associated embodied energy in FLW. The authors estimated that 2.1 billion GJ, representing at least 2
percent of total energy consumption in the U.S. in 2007, was associated with edible FLW from the consumer and
retail stages of the supply chain each year. Birney et al. (2017) updated this analysis with more recent FLW
estimates29, maintaining a similar scope, to produce an estimate of 2.5 billion GJ.
More recently, two studies examined the embodied energy use in FLW from all stages of the supply chain. Read
et al. (2020) estimated embodied energy in their analysis using EEIO models. The authors estimated 2.0 billion
GJ per person annually for all FLW. Companion studies by Pagani et al. (2020) and Vittuari et al. (2020) took a
more detailed bottom-up approach to estimating energy use associated with U.S. FLW. Using their own estimates
of FLW rates, the authors examined energy use between 2004 and 2015, using data from USDA, EIA, U.S.
Geological Survey (USGS), EPA, and others to approximate FLW mass and energy use at each stage of the food
system. The authors estimated that an average of 11.88 billion GJ of energy was used in the food system
annually between 2004 and 2015, of which 2.4 billion GJ (17 percent) was embodied in FLW. Compared to Read
et al. (2020), the only other study with a similarly comprehensive scope, Pagani et al. (2020) and Vittuari et al.
(2020) provide a greater level of detail, descriptiveness, and transparency in methodology, and thus policymakers
may find this estimate the most useful.
Notably, Pagani et al. (2020) and Vittuari et al. (2020) knowingly underestimated the energy use of FLW, as the
authors excluded the loss and waste of imported foods from their analysis due to the lack of reliable data on the
energy embodied in foreign production and international transport. One-fifth of the U.S. food supply is imported,
and almost one-half of imported foods are fruits and vegetables (USDA, 2019f, 2016) which are lost and wasted
at a relatively high rate (Buzby et al., 2014; Gustavsson et al., 2011), so impacts likely would have been
substantially higher if the authors had included loss and waste of imported foods.
Unexpectedly, the estimates from Cuellar and Webber (2010) and Birney et al. (2017), which represent only retail
and consumption FLW are similar to estimates from Read et al. (2020), Pagani et al. (2020), and Vittuari et al.
(2020), which encompass FLW from all stages of the supply chain. One would expect the latter estimates to be
larger than the former. This can be partly explained by the exclusion of imported foods by Pagani et al. (2020) and
Vittuari et al. (2020), potentially keeping their estimate roughly 20 percent lower than would be expected.
Taking all the reviewed modeling variables into account, the Pagani et al. (2020) and Vittuari et al. (2020)
estimate of 2.395 billion GJ (664 billion kWh), or 7.7 GJ (2,140 kWh) per person, annually is likely the most
precise current estimate of wasted energy inputs embodied in U.S. FLW, even though it excludes the loss and
waste of imported foods. For context, this finding suggests that FLW accounts for 2 percent of total U.S. energy
consumption and embodies enough energy to power approximately 56 million U.S. homes for a year (U.S. EPA,
2021b; EIA, 2020)30.
29 Cuellar and Webber (2010) relied on 1995 USDA FLW data, while Birney et al. (2017) used 2010 data from the same
source.
30 Based upon 2019 usage levels of approximately 42.8 GJ/year
Chapter 4. Environmental Footprint of U.S. Food Waste ^h 44
-------
Scope of
Energy Use
Energy Use
Factors
Pagarii et al. (2020);
Vittuari et al. (2020)
7.7•
U.S.
Analyzed the energy used in each part of the food
supply chain. Excludes all exports and imports.
Used EPA's USEEIO model used to estimate energy
use associated with FLW within each stage of the food
supply chain.
Extrapolated the per person energy use from Cuellar
and Webber (2010) and applied updated FLW
estimates.
Cuellar and Webber
(2010)
EssI
U.S.
Calculated the energy used to produce food from
primary production (including aquaculture and fisheries),
transportation, food processing, packaging, food
services and residential energy consumption. Relied on
one case study for food processing energy factors.
Skaf etal. (2021)
2.7*
Intl
HI
Calculated fossil fuel depletion in kg oil equivalent.
Applied the same international food production process
data from Ecolnvent for all the studied countries,
including the U.S.
• = Calculated values for tons and per person used the same population factor as the study's data year,
= personal communication with the author.
Chapter 4. Environmental Footprint of U.S. Food Waste
45
-------
Energy Use, By Supply Chain Stage and Food Category
The companion studies by Vittuari et al. (2020) and Pagani et al. (2020) described above provide detailed
estimates of the embodied energy of FLW, by supply chain stage and by food category. As shown in Figure 4-6,
the authors found that the consumption stage of the supply chain embodied the largest amount of wasted energy
from FLW in the food system—72 percent—at 1,723 million GJ. Within the consumption stage, the contribution of
at-home FLW (1,101 million GJ) exceeded that of away-from-home FLW (622 million GJ). Even though the
consumption stage accounts for a large share of FLW, its contribution to energy use is still outsized (47 percent of
FLW but 72 percent of energy use associated with FLW31). The next largest contribution was from the retail
sector. Together the consumption and retail stages account for 90 percent of energy use associated with FLW.
Of food categories examined, all categories entailed more energy use downstream than upstream on-farm or
during processing. Animal products (including meat, milk, eggs, and fish) embodied the largest amount of wasted
energy at 1418 million GJ (60 percent of the total wasted energy, despite representing only 34 percent of FLW).
Among individual food categories, meat resulted in the largest cumulative embodied energy loss (629 million GJ
or 26 percent of the total).
Pagani et al. (2020) and Vittuari et al. (2020) found that, in the upstream stages of the supply chain (primary
production and processing) each kilogram of FLW carries a burden of 10 to 40 MJ (vegetal32 products) or 30 to 75
MJ (animal products). In the downstream stages, the burden of each kilogram of FLW is 20-60 MJ (vegetal
products) or 30 to 110 MJ (animal products).
Prior to this study, Cuellar and Webber (2010) had examined the embodied energy loss of FLW by food category
by pairing food categories with mass data and energy intensities at each stage of the food supply chain, then
multiplying each category by FLW rates. Birney et al. (2017) later updated this analysis, using the same energy
use intensities. These studies highlighted dairy, meat, vegetables, and poultry and fish as substantial contributors
to the embodied energy of FLW. While categories like meat were the most energy-intensive to produce, the
contribution of other categories such as vegetables were driven by the large amount wasted.
Meat
Dairy & Eggs
Fish
Grains
Vegetables
Fruit
Sweetners
Beverages
Oilseeds & Soybeans
200 400 600
Total embodied energy waste
(million gigajoules)
0% 50% 100%
Distribution of wasted embodied
energy by supply chain stage
FIGURE 4-6. EMBODIED ENERGY OF U.S. FLW
Data Source: Pagani et al. (2020); Vittuari et al. (2020). Does not include energy use from packaging.
31 According to data in Pagani et al. (2020) and Vittuari et al. (2020)
32 The term "vegetal" here includes all food categories except animal products.
Chapter 4. Environmental Footprint of U.S. Food Waste
46
-------
4.6 Greenhouse Gas Emissions
Increasing levels of carbon dioxide and other greenhouse gases in our atmosphere, caused by human activities,
are contributing to changing earth's climate - rising temperatures, changes in precipitation, and more extreme
climate events. The food system is a major contributor to anthropogenic GHG emissions. In the U.S., agriculture
(primary production) is responsible for 10 percent of total domestic GHG emissions, not accounting for associated
emissions from land use and land use change (U.S. EPA, 2021c). Including the carbon emissions impacts from
agriculture driven land use and land use change, the North American food system accounts for 25 percent of total
North American GHG emissions (Crippa et al., 2021).
The ten studies included in Tables 4-7 and 4-8 estimate the GHG emissions associated with food that was
ultimately lost and wasted. None of the studies attempted to quantify the GHG emissions tied to land use and land
use change from FLW, likely because it typically occurs outside the U.S.33 In keeping with the scope of this report,
emissions from landfills or other food waste management methods are not considered. Table 4-8 summarizes the
results.
Methane and Nitrous Oxide Emissions
Two studies (Chen et al., 2020; Hi2 emissions (CH4 and N2O)
from primary production, thus producing the lowest estimates. The results of these studies are summarized in
Table 4-7. The differences between the two estimates are not driven by the differences in scope of the studies or
the FLW estimates underpinning the studies (both of which would predict Hi2 emissions of FLW.35
TABLE 4-7. ANNUAL METHANE AND NITROUS OXIDE EMISSIONS ASSOCIATED WITH U.S. FLW
Source
Scope of FLW
GHG Emissions - CH4 and N2O emissions
Total
(million MTCC^e)
Per Person Scope of GHG Emission
(kg CC>2e) Emissions Factors
Hi? et al. (2016)
Si
124
U.S.
Estimated emissions using country-level data on
agricultural emissions from FAOSTAT.
Chen et al. (2020)
in
167
Intl
Determined emissions by matching recently available
global average characterization factors per food group
(i.e., cropland use [g CCbe/g]) to product resolution.
= calculated value
33 While food demand and consumption in the U.S. rises with a growing population, agricultural land use has remained
relatively stable in the U.S. since the 1960's (USDA, 2017).
34 Hig et al. (2016) calculated 1050 daily calories per capita from all sectors except primary production, while Chen et al.
(2020) estimated 709 daily calories per capita from the consumption sector alone.
35 Hig et al. (2016) did not provide the ratio of vegetal to animal products used in the analysis.
Chapter 4. Environmental Footprint of U.S. Food Waste
47
-------
All Greenhouse Gas Emissions
The other eight studies evaluated all GHG emissions (including CO2 and non-C02 emissions). Read et al. (2020)
and ReFED (2021a) provide the broadest estimates, considering GHG emissions associated with FLWfrom all
supply chain stages and including GHG emissions from all stages of the supply chain (including the consumption
stage, which all other studies omitted). The six remaining studies (Guo et al., 2020; Birney et al., 2017; Heller and
Keoleian, 2015; Venkat, 2012) considered a subset of food supply chain stages when calculating FLW and/or
GHG emissions. Guo et al. (2020) also included emissions from international transportation, which no other study
did.
The highest total estimates were from ReFED (2021a) followed by Guo et al. (2020). ReFED (2021a) quantified
the GHG emissions related to FLW for the whole food supply chain, including imports and transportation.
However, their GHG emissions value during primary production only covers fruits and vegetables. Guo et al.
(2020) evaluated emissions from primary production as well as those from international transportation of food and
food products. The study accounted for FLWfrom all stages of the supply chain. The authors estimated that 222
million MTCC^e were associated with the primary production and international transportation of edible and
inedible FLW. Surprisingly, international transportation accounted for only 3 percent of the total emissions
estimate.
Heller and Keoleian (2015), followed by Birney et al. (2017), calculated GHG emissions associated with a smaller
scope of FLW but a larger scope of GHG emissions (i.e., from more supply chain stages) than Guo et al. (2020),
making comparisons difficult. Both Heller and Keoleian (2015) and Birney et al. (2017) relied on GHG emissions
factors for each food category based on a meta-analysis of LCAs of food production. The studies attribute a
higher emissions intensity to animal product categories and oils than did Guo et al. (2020). Birney et al. (2017)
estimate being higher than that of Heller and Keoleian (2015) may be partly attributable to more recent, higher
estimates of FLW, including 40 percent higher estimates of fruit and vegetable FLW. Skaf et al. (2021) is the most
recent study included, relying on food production data within Ecolnvent. Their calculated GHG emissions value
falls squarely between Birney et al, (2017) and Heller and Keoleian (2015) even though it only examines
consumer food waste.
Covering more stages of FLW but similar stages of GHG emissions to Heller and Keoleian (2015) and Birney et
al. (2017) above, Venkat (2012) used a proprietary database of LCAs and life cycle impact data of food products
to calculate a much lower GHG footprint of edible FLW. This is the lowest estimate of the studies that examined
all GHG emissions. This can be explained by Venkat using the smallest estimate of FLW (180 kg per person) and
the lowest GHG emission intensities for meat, poultry, and eggs, which comprise the greatest portion of the FLW
GHG footprint.
Looking at the studies that included all GHG emissions, only Read et al. (2020) considers FLW and GHG
emissions from all stages of the supply chain, encompassing a larger scope than the other studies - howeverthe
study does not present the highest estimate.
There are advantages and disadvantages to each type of methodology. For example, while EEIO models (like the
one used by Read et al. (2020) can provide coarse estimates due to their economy-wide view, rolling up individual
food product LCAs (akin to methodology of Birney et al. (2017) and Heller and Keoleian (2015)) can possibly
exaggerate data uncertainties and assumptions. The choice of FLW estimate and GHG emissions factors in all
the studies influence final estimates. More research on GHG emissions associated with FLW is warranted. For
now, the authors of this paper recommend use of the possibly conservative estimate from Read et al., with the
understanding that other studies (except Venkat, 2012) indicate it may be an underestimate.
Taking all the reviewed modeling variables into account, Read et al. (2020) provides the most comprehensive
estimate of GHG associated with FLW, at 170 million MTCC>2e, or 539 kg C02e per person, annually. This is
equivalent to more than the emissions of 42 coal-fired power plants or 36 million passenger vehicles each year
(U.S. EPA, 2021a). However, other studies with smaller scopes present consistently larger estimates, warranting
further study.
Chapter 4. Environmental Footprint of U.S. Food Waste
48
-------
TABLE 4-8. ANNUAL GHG EMISSIONS ASSOCIATED WITH U.S. FLW
GHG Emissions - All
Source
Scope of FLW
Total
(million MTC02e)
Per Person
(kg C02e)
Scope of GHG
Emissions
Emission
Factors
ReFED (2021a)
rm
822"
ppi
Varies
Calculated the life cycling impacts, using Quantis'
database, for 44 common food items to represent the US
food market from the farm to residential stage.
Read et al. (2020)
539
U.S.
Used EPA's USEEIO model to estimate GHG emissions
associated with FLW.
Guo et al. (2020)
683
FAO
Included emissions from international transportation.
Used regional and food specific emission factors from
LCAs for primary production and FAO detailed trade
matrix data and per-km emission factors for
transportation.
Venkat (2012)
368
Varies
Estimates emissions for each food category based on an
LCA framework.
Birney et al. (2017)
Used Heller and Keoleian (2015) emissions data.
Heller and Keoleian
(2015)
latlXH
511
Varies
Emission factors based on meta-analysis of published
LCA values for various food types from both domestic
and international studies.
EPA WARM (2019)a
Skaf et al. (2021)
650-
U.S.
Uses streamlined iifecycle emission factors for FLW
including five primary food categories from primary
production to retail, including transportation.
527
Varies
Used LCA data for food production from Ecolnvent and
applied the ReCiPe midpoint method.
* = calculated value based on WARM emission factors applied to Buzby et al. (2014) FLW data.
= personal communication with the author
a EPA WARM is a tool that can calculate estimated Iifecycle GHG emissions associated with food waste. The FLW values by food category
from Buzby et al. (2014) were entered into WARM to develop these estimates.
Chapter 4. Environmental Footprint of U.S. Food Waste
49
-------
EPA's Waste Reduction Model (WARM)
EPA's Waste Reduction Model (WARM) is a tool to compare the GHG emissions associated with different
waste management pathways (e.g., landfill or composting) for many different material types, including food.
WARM can also be used to see the GHG emissions associated with consumption stage FLW of beef, poultry,
grains, bread, fruits and vegetables, and dairy products.
WARM employs a streamlined life cycle analysis, providing information on GHG emissions from the:
primary production of food;
transport of materials from the production or processing facility to the retail/distribution
point;
manufacture and application of agricultural fertilizers;
management of livestock manure;
enteric fermentation resulting from livestock; and
fugitive emissions of refrigerants used during refrigerated transport and storage.
As seen in Figure 4-7, the resulting emission factors from primary production to consumption vary by food
product from 33 MTC02e/metric ton to 0.5 MTC02e/metric ton for beef and fruits and vegetables,
respectively. Applying the WARM GHG emission factors (excluding disposal) to the 2010 FLW estimate from
Buzby et al. (2014) results in an approximate GHG footprint of 650 kg of C02e per a person. This falls within
the range of 368 to 683 kg C02e per person (Guo et al., 2020; Venkat, 2012) from the estimates discussed
above that examined CO2 and non-CC>2 emissions from food waste.
Beef
Poultry
Dairy Products
Fruits & Vegetables |
I
I
I I
J
I
I
Grains |
Source
(in vertical order)
Venkat (2012)
Heller & Keoleian (2015)
Guo et al. (2020)
U.S. EPA WARM v.15
10 15 20 25
kg C02e / kg food
30
35
FIGURE 4-6. GHG EMISSIONS INTENSITIES, BY FOOD CATEGORY
Data from Guo et al. (2020) obtained from personal communication with X. Guo (March 23, 2021).
Chapter 4. Environmental Footprint of U.S. Food Waste ^h 50
-------
GHG Emissions, By Supply Chain Stage and Food Category
Several insights about the distribution of FLW-associated GHG emissions along the supply chain are available in
the literature, most notably that primary production is a far greater contributor to GHG emissions than
transportation. Guo et al. (2020) examined GHG emissions from primary production and international
transportation, and found international transportation accounted for only 3 percent of total emissions. Venkat
(2012) assessed GHG emissions from primary production through retail (i.e., including emissions from domestic
transportation and excluding emissions during the consumption stage), finding that primary production and food
processing36 accounted for 80 percent of the GHG emissions, followed by 14 percent from distribution and retail,
and 6 percent from packaging. Distribution and retail would include both domestic transportation and energy use
and refrigerant-related emissions at retail outlets. Neither study considered consumption stage emissions, which
are likely to be significant based upon the energy use estimates presented in Section 4.5.
Looking at specific food groups' contributions to the GHG emission footprint of FLW, animal products, particularly
ruminant-based FLW (i.e., dairy and beef) result in the majority of emissions. For example, Heller and Keoleian
(2015) examined GHG emissions (excluding emissions from the consumption stage) from food lost or wasted at
the retail and consumption stages. The authors found that the beef, veal, and iamb category accounted for the
greatest GHG emissions, followed by dairy products (other than fluid milk) and pork. Together animal products
(beef, veal, and lamb; milk and other dairy products; pork; poultry; fish and seafood; eggs) accounted for 73
percent of GHG emissions from retail and consumer FLW, while accounting for only 33 percent FLW by weight
and 23 percent FLW by calories. Guo et al. (2020) similarly found that for consumption stage FLW in the NAO
region, beef represented 44 percent of cradle-to-consumer GHG emissions associated with FLW; together with
dairy, all ruminant FLW accounted for 60 percent of GHG emissions associated with FLW.
Figure 4-8 shows the contribution of each food category to the GHG footprint of FLW by displaying data from the
four available studies. Figure 4-6 compares the GHG emissions intensities (i.e., the emissions per unit of food) of
a few specific food categories, based upon data from three of these same studies, demonstrating that beef has
the highest GHG emissions intensity.
Meat, Poultry, Eggs
Dairy
Grains
Fish Fat
Sweeteners
Heller &
Keoleian (2015)
FLW
-ir~
n
ii ii ii i
I
\^m
Guo et al. (2020)
FLW
¦
¦
1 ¦
Bimey et al. (2017)
FLW
Venkat(2012)
FLW
0%
20%
40%
60%
80%
100%
FIGURE 4-8. COMPARISON OF GHG FOOTPRINT TO FLW COMPOSITION
U.S. data from Guo et al. (2020) obtained through personal communication with the lead author.
36 Primary production and processing are combined in Venkat (2012) published data, as are distribution and retail.
Chapter 4. Environmental Footprint of U.S. Food Waste
51
-------
4.7 Summary of Environmental Footprint
Despite differences in study design, methodologies, data sets, time periods, and other factors, most estimates of
the environmental impacts associated with food that is ultimately lost or wasted show general agreement once
these factors are taken into consideration. Table 4-9 presents selected estimates of the annual cradle-to-
consumer environmental impacts of FLWin the U.S. in absolute and per person terms. Figure 4-10 displays this
data as a percentage of the environmental footprint of the entire U.S. cradle-to-consumer food system.
In general, studies show that roughly a third of the U.S. food supply is lost or wasted, and FLW accounts for
roughly one-third of the inputs and environmental impacts of the cradle-to-consumer food system. For example,
Birney et al. (2017) concluded that food lost or wasted during the retail and consumption stages uses
approximately one-third of all resources of the food system,37 as might be expected if FLW from along the entire
chain had been included. Kummu et al. (2012) examined FLW along the entire supply chain, also finding one-third
of resources to be associated with FLW.38 Another study that did included FLW all along the supply chain,
however, produced a lower estimate. Read et al. (2020) EEIO analysis found that approximately 16 to 18 percent
of the total environmental impact39 of the U.S. cradle-to-consumer food system is associated with food that is
ultimately lost or wasted. Other studies produced estimates higher than one-third. For example, Toth and Dou
(2016) estimated more than 40 percent of irrigation water and cropland were associated with FLW.
While a disparity between the amount of food lost or wasted and the portion of the food system's inputs and
environmental impacts may be due in part to the point on the supply chain at which the food is lost or wasted
and/or the mix of food categories lost or wasted, it could also imply that the relationship between FLW and food
system impacts is more complex than it appears. Also, the differences in estimates above are difficult to compare,
due to differences in scope and methodologies. Certainly it is clear that downstream FLW, especially at the
consumption stage (i.e., at restaurants and at home), is more of an environmental burden that FLW further
upstream, per unit of food, as the inputs and impacts accumulate as food moves along the supply chain.
While the estimates presented in Table 4-9 were chosen largely based upon their comprehensive scope, many
credible methodologies presented in this report would produce higher estimates if their results were extrapolated
to cover FLW along the whole supply chain. The estimates of Read et al. (2020) in particular should be
considered conservative.
37 Birney et al. (2017) includes energy, blue water, green water, GHGs, agricultural land, and fertilizer. The authors found FLW
accounted for 5% of energy use, 34% of blue water use, 34% of GHG emissions, 31% of land use, and 35% of fertilizer use
related to an individual's food-related resource consumption, i.e. their footprint.
38 Kummu et al. (2012) included the use of blue water (35%), cropland (31%) and fertilizers (30%) in this estimate.
39 Read et al. (2020) includes energy use, eutrophication potential, GHG warming potential, land use and water use (blue
water withdrawals).
Chapter 4. Environmental Footprint of U.S. Food Waste
52
-------
TABLE 4-9. SUMMARY OF THE ANNUAL CRADLE-TO-CONSUMER ENVIRONMENTAL FOOTPRINT OF U.S. FLW
Environmental Footprint
Environmental
Impact
Total
(Standard Units)
Per Person
Percentage of U.S.
Cradle-to-Consumer
Food System
Footprint
Percentage of U.S.
Footprint
Source
Scope of FLW
%r
Land
Use
560,000 km2 •
(140 million acres)
1,800 m2 f
(19,000 sq ft)
16% of agricultural
land •
-
Read et al. (2020)
S&
Water
22 trillion L *
71,000 L f
17% of freshwater
5%
Read et al. (2020)
Use a
(5.9 trillion gallons)
(19,000 gallons)
used •
| )©)#H
10
Pesticide
Application
350 million kg b
(780 million pounds)
1 kg •
(2.5 pounds)
-
-
Conrad et al. (2018)
H
@
Fertilizer
Application
6,350 million kg *
(14 billion pounds)
20.2 kg
(44.5 pounds)
42% of total fertilizers
used
-
Toth and Dou (2016)
rpf
Energy
Use
2,400 million GJ
(664 billion kWh)
7.7 GJ *
(2,140 kWh)
20% of energy used
2%
Pagani et al. (2020);
Vittuari et al. (2020)
GHG
©
GHG
Emissions
170 million MTCCbe*
f
540 kg CCbe
16% of GHG
emissions •
2%
Read et al. (2020)
• = calculated
f= personal communication with author
a Blue water use.
b Accounts for only consumer FLW
Chapter 4. Environmental Footprint of U.S. Food Waste
53
-------
Farm-to-Kitchen
Environmental Footprint of
U.S. Food Loss and Waste
(excluding impacts of waste management,
such as landfill methane emissions)
GHG emissions of
42 coal-fired power plants
Rertilize^^j
The amount of fertilizer used
to grow all plant-based foods
Enough water and energy to supply
more than 50 million homes
An area of agricultural land
equal to California and New York
FIGURE 4-9. ANNUAL CRADLE-TO-CONSUMER ENVIRONMENTAL FOOTPRINT OF U.S. FLW
This figure depicts the annual environmental footprint of producing, storing, processing, packaging, distributing, and marketing food
that is ultimately lost or wasted in the United States. Data Source: U.S. EPA (2021a); USCB (2021); Pagani et al. (2020); Read et al.
(2020); U.S. DoE (2020); Vittuari et al. (2020); U.S. EPA (2018); Toth and Dou (2016)
Chapter 4. Environmental Footprint of U.S. Food Waste
54
-------
For policymakers seeking credible estimates of the environmental footprint of retail and consumption stage FLW
only, in line with the UN SDG Target 12.3 and the EPA and USDA goal, Birney et al, (2017) is the most useful
resource. The authors estimate 325,500 km2 (80 million acres) agricultural land, 17 trillion L (4 trillion gallons)
blue water, 123 trillion L (32 trillion gallons) green water, 5,266 million kg (12 billion pounds) fertilizer, 2.5 billion
GJ (694 billion kWh) energy, and 208 million MTCCbe GHG emissions are associated with retail and consumption
stage FLW annually, from cradle-to-consumer.
In addition to demonstrating the resource inputs and environmental impacts associated with U.S. FLW, the
studies presented provide evidence of the factors (food categories and food supply chain stages) driving these
estimates. This can provide policymakers with clues as to how to maximize environmental benefits of FLW
reduction initiatives. For example, while the use of land, pesticides, fertilizers, and water chiefly occur during
primary production, energy use and GHG emissions occur all along the supply chain. Studies illuminated that the
consumption and distribution stages account for the greatest energy use, while the primary production stage
accounts for the greatest GHG emissions.
The type of food lost or wasted also has a significant effect on the environmental footprint of FLW, and
policymakers may prioritize interventions related to categories of FLW with the largest environmental impacts.
Given the predominance of animal products and fruits and vegetables for each input and environmental impact,
policymakers may want to consider FLW initiatives targeting these food types. Animal products40 are responsible
for more than half of the land and energy used, and GHGs emitted from FLW (Conrad et al., 2018; Birney et al.,
2017; Toth and Dou, 2016) and accounted for the largest share of fertilizer and water use for irrigation (Conrad et
al., 2018). Fruits and vegetables were also substantial users of inputs, ranking second behind animal products in
many categories. Fruit also accounted for the greatest pesticide application, followed by animal products (Conrad
et al., 2018). However, fruits and vegetables comprise a much larger share of U.S. FLW than animal products,
demonstrating the outsized impact of the loss and waste of animal products.
FLW Consumed by Humans
If
Land Use
140
million
acres
P
Water Use
5.9
trillion
gallons
a
14
Fertilizer Use
billion pounds
Sum N, P205, & K20
664
Energy Use
billion kWh
GHG
million
GHG Emissions
MTC02e
FIGURE 4-10. ANNUAL ENVIRONMENTAL FOOTPRINT OF
U.S. FARM TO KITCHEN FOOD SUPPLY CHAIN
Data Source: Read et al. (2020) (land, water- blue water, and GHG) - noting that these are calculated values based from per capita data
received in personal communication; Toth and Dou (2016) (fertilizer); Pagani etal. (2020) & Vittuari etal. (2020) (energy)
40 Including the feed crops that support animal production
Chapter 4. Environmental Footprint of U.S. Food Waste
55
-------
CHAPTER 5.
Environmental Benefits of Reducing
U.S. Food Loss and Waste
Given the significant environmental impacts of FLW, halving FLW-
as the U.S. aims to do - could meaningfully reduce the resource use
and environmental impacts of the U.S. food system. This chapter
examines the potential environmental benefits of halving FLW in the
United States.
Building upon the analyses of the environmental footprint of FLW, as
presented in Chapter 4, researchers have estimated the potential
"savings" (i.e., avoided resource use and environmental impacts)
that could be achieved by reducing U.S. FLW. When calculating
savings, researchers must consider the supply chain stage at which
the reduction was achieved and the category of food in which waste
was prevented. The environmental benefits presented in this chapter
can only be achieved through the prevention (i.e., source reduction)
of food waste. Recycling food waste will not achieve these
benefits.41
The methodologies upon which all these estimates are built assume
that decreases in demand/consumption (from reducing FLW) will
result in equivalent decreases in production. However, economic
factors like rebound effects can impede reductions in production,
and thus these estimates of environmental savings should be
considered the upper bounds of savings that could be achieved.
5.1 Environmental Benefits,
Relative to Current Footprint
Six recent studies estimated the percentage of the U.S. cradle-to-
consumer food system's environmental footprint that could be saved
(or avoided) if the U.S. reduced FLW. The studies considered inputs
and environmental impacts similar to those discussed in Chapter 4 -
the use of land for agriculture; use of blue water, fertilizer, and
energy; and GHG emissions associated with FLW. Table 5-1
compares the methodologies and results of the six studies.
Four of the studies modeled halving U.S. FLW, while the remaining
two studies modeled slightly greater reductions. ReFED (2021a)
modeled 56 percent reduction, and Kummu et al. (2012) modeled a
roughly 63 percent reduction. Kummu et al. (2012) derived its target
by modeling a scenario where each of seven world regions
(including NAO) achieved the lowest current FLW rate (among all
seven regions) for each food category in each supply chain stage
globally.
KEY FINDINGS
¦ Halving U.S. FLW could
achieve the following
annual savings:
¦ Agricultural land: More
than 300,000 km2 (75 million
acres)
O Blue water:
12 trillion L (3.2 trillion
gallons)
O Fertilizer:
Nearly 290,000 metric
tons (640 million
pounds) bioavailable
nitrogen
O Energy:
940 million GJ (262
trillion kWh)
O GHG emissions:
92 million MTCC>2e
¦ Halving FLW in
households, restaurants,
and the food processing
sector will have the
greatest environmental
benefits. Halving the FLW
in retail and institutional
food service (schools) will
have minimal
environmental benefits.
- Reducing loss and waste
of meats, cereals, and
fresh fruits and vegetables
will have the greatest
environmental benefits,
among food categories.
41 EPA's forthcoming companion report (The Environmental Impacts of U.S. Food Waste: Part 2) will compare the
environmental footprint of food waste prevention to that of food waste management pathways, such as landfills, combustion,
composting, and anerobic digestion.
Chapter 5. Benefits of Reducing U.S. Food Waste
56
-------
Other than the level of reduction modeled, the key difference among the six studies' methodologies is how they
reported results. Four of the studies ((Read et al., 2020; Wood et al., 2019; Springmann et al., 2018; Jalava et al.,
2016; Kummu et al., 2012) estimate savings associated with reducing FLW relative to the current environmental
footprint of the food system, while Springmann et al. (2018) estimates savings relative to a future business-as-
usual (BAU) scenario. ReFED (2021a) reported annual reductions only in absolute terms.
Examining estimates from the first four studies (Read et al., 2020; Wood et al., 2019; Springmann et al., 2018;
Jalava et al., 2016; Kummu et al., 2012), there is general agreement among three of the studies—Kummu et al.
(2012); Wood et al. (2019); and Jalava et al. (2016)—about the magnitude of environmental benefits that could be
achieved by reducing U.S. FLW. The three studies estimate reductions ranging from 13 to 16 percent across the
inputs and environmental impacts measured by more than one of the studies - use of land for agriculture, use of
blue water, fertilizer, and energy; and GHG emissions.42 Uniquely, Wood et al. (2019) estimated potential
reductions in ammonia emissions from fertilizer use (14 percent) and Jalava et al. (2016) estimated potential
savings of green water (12 percent). No other studies addressed these inputs, but these estimates present
roughly similar magnitudes of savings to the other resources and impacts in the three studies.
The main differences among the three studies' methodologies include scope and treatment of imports. Wood et
al. (2019) focused exclusively on halving FLW from retail and consumption stages, while all other studies in this
chapter modeled reducing FLW all along the supply chain. While most studies presented in this chapter (and this
report) assumed the U.S. food supply was domestically produced when calculating environmental impacts, Jalava
et al. (2016) attempted to improve accuracy by applying the global (rather than U.S.) average water use factor to
imported foods that were lost or wasted.
The estimates presented in the above studies are in line with modest expectations. If roughly one-third43 of the
U.S. food supply is lost or wasted (see Chapter 3), and FLW accounts for roughly one-third44 of the inputs and
environmental impacts of the U.S. food supply (see Chapter 4), then halving FLW would be expected to achieve
savings of around one-sixth (i.e., one half of one third, or 17 percent) of the current food system's inputs and
environmental impacts. Thus, the results of these three studies, at 12 to 16 percent, are not surprising.
Differences between the makeup of FLW (e.g., food categories and supply chain stages at which food was lost or
wasted) and the makeup of the overall food supply could account for the differences.
However, the percentage estimates of Read et al. (2020) are consistently lower than those of the other three
studies, ranging from 9 to 10 percent. As described in Chapter 4, Read et al. (2020)'s use of an EEIO model may
explain this difference; however, EEIO models' inclusion of intermediate inputs,45 in addition to primary inputs,
would be anticipated to increase both the inputs and savings, but the authors' results were lower than, not higher
than, many other studies, once differences in scope are considered.
Looking at specific environmental measures, two other methodological differences emerge between Read et al.
(2020) and other studies. First, while many other studies in this chapter and the previous chapter measure the use of
fertilizer, Read et al. (2020) modeled eutrophication potential (i.e., nitrogen releases due to fertilizer use) by
combining nitrogen fertilizer application data with published factors and models of loss. While this metric is
potentially more useful than simply estimating fertilizer use, as it estimates the environmental impact, rather than just
the input, the results are not directly comparable to that of the other presented studies. Read et al. (2020) did not
publish an estimate of fertilizer use. Read et al. (2020) estimates a 10 percent reduction in eutrophication potential
from halving U.S. FLW, while the others estimate a 14 to 16 percent reduction in nitrogen and phosphorous fertilizer
application.
42 The percentages are in relation to each individual study's baseline which is influenced by the breadth of stages included and
methodology. For example, Wood et al. (2019) estimates an environmental savings of 1.2 trillion L by halving FLW which
represents a 14% reduction compared to her baseline. Read et al. (2020) estimates an environmental savings of 12.2 trillion L
by halving FLW which represents a 9% reduction compared to his baseline. Wood was only examining the environmental
impacts and potential savings of FLW reductions from the retail and consumption stages, whereas Read was looking across
the full supply chain.
43 Estimates range from 25 to 45 percent, when measured by weight or calories. See Table 3-1.
44 Estimates vary by study and by input or environmental impact. Results from Read et al. (2020) excluded. Estimates reported
in the literature as percentages of the total cradle-to-consumer food system include: 29 percent of GHG emissions (Venkat,
2012), approximately one third of blue water to produce crops and livestock and 30 percent agricultural land (Birney et al.,
2017), and 42 percent of cultivated cropland and 44 percent of water used for irrigation (Toth and Dou, 2016). None of these
studies include FLW from all supply chain stages
45 For example, EEIO models include both the freshwater used in primary production, along with the additional inputs used to
deliver that water, whereas the methodologies of Jalava et al. (2016), Kummu et al. (2012), and Wood et al. (2019) quantify
just the first order inputs.
Chapter 5. Benefits of Reducing U.S. Food Waste
57
-------
TABLE 5-1. MAXIMUM ENVIRONMENTAL BENEFITS OF HALVING U.S. FLW
Environmental
Impact
Water
(million L)
Fertilizer
(million kg)
Energy
(million GJ)
GHG
(million MTCC>2e)
Environmental Savings
Relative to Current Food System Footprint
(as determined by source)
Relative to 2050
BAU Scenario
Jalava
et al. (2016)
Kummu et al. Read ReFED
(2012) etal. (2020) (2021)
Wood
et al. (2020)
Springmann
et al. (2018)
H
WtonM&m B—H
[2
82,800
300,000*
427,000 209,000
J- ~
6,400,000
~ ^
12,000,000*
15,000,000
1,200,000
1,500 NPK [eutrophication indicator]
J- ~
112
940
92 *
75
577
87.5
24,000,000
2,930 N, 402 P
41
I estimates represent results from a 50% FLW reduction, except Kummu et al., 2012, (63%) and ReFED, 2021a, (56%).
Chapter 5. Benefits of Reducing U.S. Food Waste
58
-------
In addition, Read et al. (2020) estimates 6 percent lower energy savings than Wood et al. (2019) (9 and 15 percent,
respectively), likely due to differences in the scope of energy use included in the two studies. Read et al. (2020)
included energy use all along the cradle-to-consumer supply chain, while Wood et al. (2019) excluded energy use
during the consumption stage. Energy use in the consumption stage of the food supply chain is chiefly for
refrigeration and cooking (Vittuari et al., 2020), and it is unclearto the authors of this paper how the prevention of
FLWwould significantly reduce this type of energy use. If the food that was ultimately wasted was never purchased,
for example, a household's refrigerator would still be running and it may, in fact, use more energy to cool a less full
refrigerator. If Read et al.'s (2020) model accounted for this dynamic, it could be anticipated to produce lower energy
savings in the consumption stage than in other stages. Given that consumption stage energy use accounts for the
majority of the supply chain's energy use (Pagani et al., 2020; Vittuari et al., 2020), Wood et al. (2019) would be
expected to project a larger decrease in energy use than Read et al. (2020).
ReFED (2021a) does not provide percentage values (nor do their estimates of savings rely on estimates of
impacts of the total food supply chain); however, their estimated savings of water and GHG emissions are in line
with the other studies. Their GHG emissions savings may be understated because the emissions from meat and
dairy were not included within primary production.
5.2 Environmental Benefits, Relative to Future Footprint
Four of the five abovementioned studies (all but ReFED (2021a)) measured savings relative to the current
environmental footprint of the cradle-to-consumer food system; however, the sixth study (Springmann et al., 2018)
considered projected changes in food production and consumption between 2010 and 2050 (using the IMPACT
model46) when estimating benefits of halving FLW. This is important (especially for the projections of global
environmental benefits in the next chapter) as global food production and consumption are expected to change
substantially in coming decades due to socioeconomic factors, such as population and income growth, and
environmental pressures will increase as a result.
By halving FLW, Springmann et al. (2018) projects the U.S. could achieve reductions of 14 to 16 percent from a
2050 BAU food system footprint, with regard to agricultural land use, water use, and nitrogen application; and a 9
percent reduction in non-CC>2 GHG emissions from primary production. Note that Springmann et al. (2018) looked
exclusively at CH4 and N2O emissions during primary production, rather than at all GHG emissions, during the
entire cradle-to-consumer food system, like other studies described above (Read et al., 2020; Wood et al., 2019).
Springmann et al.'s 2050 BAU scenario includes population growth and demographic changes but does not
include any new dedicated measures to mitigate the environmental impacts of the food system, such as
technological advances or shifts to less environmentally-intensive diets. Additionally, although yields are expected
to increase by 2050, this study did not include any expected yield gains, or improvements in livestock or nitrogen
efficiencies when developing the 2050 BAU scenario.
Many of the studies presented in this chapter also evaluated other measures to move toward a more sustainable
future, including dietary shifts (toward healthier foods or towards less resource-intensive foods) and
improvements in yields and resource efficiency. These studies compared the benefits of each strategy and
evaluated combinations of strategies, pairing FLW reduction with some or all of the other strategies studied,
finding greater benefits, as would be expected.
46 IMPACT = International Model
Chapter 5. Benefits of Reducing U.S.
for Policy Analysis of Agricultural Commodities and Trade
Food Waste
59
-------
"Savings" from Halving
U.S. Food Loss and Waste
(excluding impacts of waste management,
such as landfill methane emissions)
GHG emissions
equal to
23 coal-fired
power plants
Enough water
and energy to
supply more than
20 million
homes
An area of
agricultural land
as large as
Arizona
FIGURE 5-1. MAXIMUM ENVIRONMENTAL BENEFITS OF HALVING U.S. FLW
This figure depicts the projected annual savings from halving U.S. food loss and waste. The figure examines the cradle-to-consumer food
supply chain only and thus excludes additional savings of methane emissions from landfills.
Data Source: Readetal. (2020); U.S. EPA (2021a, 2018); USCB (2021); U.S. DOE (2020)
Chapter 5, Benefits of Reducing U.S. Food Waste
60
-------
5.3 Summary of Environmental Benefits
Taking all the reviewed methodologies into account, Read et al. (2020) provides the most comprehensive estimates
of potential environmental savings from halving FLW in the U.S. However, as with the study's estimates of the total
impacts of FLW, other studies' methodologies may estimate greater reductions if they were extrapolated to cover the
same scope Read et al. (2020). Also, none of the studies presented here accounted for economic factors such as
rebound effects that may impede realization of environmental benefits, and thus should be considered each authors'
estimates of upper bound for benefits. In summary, Read et al. (2020) estimates the following annual environmental
savings (i.e., avoided inputs and environmental impacts) from halving U.S. FLW:
More than 300,000 square km2 (75 million acres) agricultural land - an area greater than the State
of Arizona;47
12 trillion L (3.2 trillion gallons) blue water - equal to the water use of 29 million American homes;48
Nearly 290,000 metric tons (640 million pounds) of bioavailable nitrogen from agricultural fertilizer
with the potential to reach a body of water, cause algal blooms and deteriorate water quality;49
940 million GJ (262 trillion kWh) energy - enough to power 21.5 million U.S. homes;50 and
92 million MTCC^e GHG - equal to the CO2 emissions from 23 coal fired power plants in a year.51
For policymakers wanting to project the potential environmental benefits of halving FLW in only the retail and
consumption stages, akin to UN SDG Target 12.3 and the EPA and USDA goal, Wood et al. (2019) may prove the
most useful resource. The authors estimate the following maximum potential savings from halving U.S. retail and
consumption stage FLW annually, from cradle-to-store: 427,000 km2 (106 million acres) agricultural land, 1,200
trillion L (317 billion gallons) blue water, 337,000 million kg (743 billion pounds) phosphorous fertilizer, 577 million GJ
(160 billion kWh) energy, and 88 million MTCC>2e.
5.4 Environmental Benefits, By Food Category and
Supply Chain Stage
In addition to modeling the total benefits of halving FLW, Read et al. (2020) modeled the environmental impacts of
halving each category of wasted food. Comparing across 13 food categories,52 Read et al. (2020) found that
reducing FLW in the meats and cereals categories resulted in the largest reductions in energy, eutrophication
potential, land use, and GHG emissions, whereas reducing FLW in the cereals and fresh fruits and vegetables
categories resulted in the largest reduction in blue water use. Wood et al. (2019) similarly identified animal
products as the key contributor to phosphorous fertilizer application and ammonia emissions (due to nitrogen
fertilizer application) associated with FLW, and the waste of fruits as the key contributor to the blue water footprint
of FLW.
Read et al. (2020) also simulated the embodied environmental impacts of halving FLW at each stage in the
supply chain, in order to identify the sectors in which reductions could provide the greatest environmental
benefits. The authors modeled all four supply chain stages, further breaking down the consumption stage into
three sectors: foodservice (restaurants), institutional food service (schools and hospitals), and households, for a
total of six stages or sectors.
47 State of Arizona land base is 72.7 million acres and is the sixth largest U.S. state
48 EPA WaterSense. "The average American family uses more than 300 gallons of water per day at home." (U.S. EPA, 2018).
49 Explanation of the eutrophication indicator taken from the EPA User Manual for the Sustainable Materials Management
Prioritization Tools, which is built on the USEEIO.
50 "In 2018, 120.3 million homes in the United States consumed 1,462 billion kilowatt-hours (kWh) of electricity (U.S. DoE,
2020). On average, each home consumed 12,146 kWh of delivered electricity (U.S. DoE, 2020)". Source: (U.S. EPA, 2021a)
51 EPA Greenhouse Gas Equivalencies Calculator (U.S. EPA, 2021a)
52 Food categories included: beverages, cereals, eggs, fresh fish and seafood, processed fish and seafood, fresh fruits and
vegetables, processed fruits and vegetables, meat, milk, oilseeds and pulses, fresh roots and tubers, processed roots, and
tubers and sugar.
Chapter 5. Benefits of Reducing U.S. Food Waste
61
-------
As shown in Figure 5-2, the authors found that halving FLW at every stage of the supply chain could reduce the
environmental footprint of the U.S. cradle-to-consumer food supply chain by 8 to 10 percent. However, the bulk of
these reductions could be achieved by halving FLW in only three of the six sectors analyzed: food processing,
restaurants, and households. For example, halving FLW in those three stages (i.e., food processing, restaurants,
and households) could reduce GHG emissions by almost 8 percent, while halving FLW in the three remaining
stages (i.e., primary production, retail, and schools/hospitals) increased the reduction by less than 1 percent.
Even when considering individual food categories there was little variation among which stages had the highest
environmental benefits or the order in which they ranked (Read et al, 2020).
Among the six sectors, the authors found that the largest reductions in GHG emissions and energy use could be
achieved by halving FLW in restaurants; the greatest reductions in agricultural land use and eutrophication
potential (due to nitrogen fertilizer application) could be achieved through halving FLW from food processing; and
the largest reduction in blue water use could be achieved by halving FLW in households. Overall, the authors
found that halving FLW at retail and schools/hospitals carried minimal environmental benefits due to the relatively
low current rate of FLW compared with other sectors (Read et al., 2020).
Water Use
Land Use
Eutrophication
Potential
Energy Use
GHG
Emissions
Households
Households
Restaurants
Processing
Processing
Households
Restaurants
Food
Processing
Households
Restaurants
Food
Processing
Households
Restaurants
Food
Processing
7.8%
3.5%
9.0%
9.8%
All Stages
3 Most Influential Stages
FIGURE 5-2. MAXIMUM ENVIRONMENTAL BENEFITS OF HALVING FLW, BY SUPPLY CHAIN STAGE
Data Source: Read et al. (2020)
Chapter 5. Benefits of Reducing U.S. Food Waste 82
-------
CHAPTER 6.
U.S. Food Loss and Waste
in Global Context
This chapter examines U.S. FLW in global context to evaluate the
U.S. contribution to this global issue and to highlight key similarities
and differences among regions and countries. This information can
guide U.S. policymakers as they set priorities and tailor FLW
reduction policies to maximize their environmental benefit, as the
most effective solutions may vary across the globe. The chapter also
provides a snapshot of the environmental benefits that could be
achieved by global achievement of the UN Sustainable Development
Goal Target 12.3 to halve FLW by 2030.
6.1 U.S. Share of Global FLW
In 2007, the U.S. was responsible for approximately 10 percent of
global edible FLW (by weight) but accounted for less than five
percent of the world's population (UN, 2020a, b; CEC, 2017; FAO,
2011). As shown in Figure 6-1, the U.S. generated the third largest
absolute amount of FLW by weight (168 million tons per year) of any
country in 2017, preceded by China and India (Guo et al., 2020).
KEY FINDINGS
Global FL W contributes 3.7
gigatons of CC>2e annually,
excluding landfill methane
emissions.
¦ Several countries, including
the U.K. and Japan, have
reported significant progress
toward halving FLW.
¦ The U.S. is responsible for 10
percent of global FLW, while
accounting for less than 5
percent of global population.
¦ The U.S. exceeds the
average per person FLW and
FLW-related GHG emissions
of high-income countries by
roughly a third.
¦ Downstream FLW and animal
product FLW comprise a
greater share of U.S. FLW
than of global FLW, thus the
environmental footprint of
each unit of U.S. FLW is
substantially larger than the
global average.
¦ Halving global FL W could
reduce cumulative global food
system greenhouse gas
emissions by 24 percent
between 2020 and 2100 (331
Gt CC>2e), relative to a
business-as-usual scenario.
Chapter 6. U.S. Food Waste in Global Context ^¦ 63
-------
-10%
U.S. Percent U.S. Percent
of Global of Global
Population Edible FLW
(by weight)
China
India
U.S.
Indonesia
Brazil
Nigeria
Russia
Mexico
Pakistan
Malaysia
0
100 200 300
Million Metric Tons
400
500
FIGURE 6-1. GLOBAL SIGNIFICANCE OF U.S. FLW
The U.S. has less than five percent of the world's population but generates approximately 10 percent of the world's food loss and
waste (FLW). In 2017, the U.S. generated the third largest amount of FLW by weight (168 million tons per year) of any country.
Data Source: UN (2020a, b); CEC (2017); FAO (2011); Guo etal. (2020)
6.2 Share of Food Supply Lost or Wasted
The FAO (2011) estimated that in 2007, approximately one-third of edible food, by weight, was lost or wasted
globally.53 Using FAO's regional data, the World Resources Institute (WRI) estimated that the 2009 average share
of total food lost or wasted, was relatively similar across all seven major geographic regions of the world,54 with 34
to 36 percent lost on average in all regions except South and Southeast Asia (26 percent) (Lipinski et al., 2013).
More recently, Guo et al. (2020) estimated 29 percent of all food (edible and inedible parts) was lost or wasted
globally in 2017.55 By comparison, the U.S. lost or wasted 35 to 36 percent of food (edible and inedible parts),
according to the most comprehensive estimates presented in Section 3.2 (ReFED, 2021a; CEC, 2017). While the
share of food supply (by weight) that is lost or wasted may be similar across many world regions, the size of each
region's food supply varies widely (even on a per person basis), limiting this metric's usefulness.
When FLW measured in calories instead of by weight, the share of total food lost or wasted is estimated to be
much higher for the NAO region (which includes the U.S.) at 42 percent, compared with 15 to 25 percent in all
other regions (Lipinski et al., 2013). In addition, the size of the per person food supply in the NAO region (4,230
calories per person per day) exceeds that of any other world region by more than 1,200 calories per person per
day (Kummu et al., 2012).56
53 FAO developed this estimate using a top-down approach that incorporated country-specific production volumes and food
balance sheets and regional level waste generation factors across the different stages of the food supply chain. It should be
noted that much of the consumption stage data for undeveloped regions were derived from limited to no primary data from
these regions.
54 Regions include: Sub-Saharan Africa; Europe (including Russia); Industrialized Asia; Latin America; North Africa and West-
Central Asia; North America and Oceania; and South and Southeast Asia.
55 Guo used FAO (2017) food balance sheets coupled with regional waste generation factors obtained from Porter et al.
(2016).
56 This study is one of the primary bases for the global goal to reduce FLW by 50 percent.
Chapter 6. U.S. Food Waste in Global Context
64
-------
A comparative analysis of the potential to reduce FLW (using best available practices and technologies) found
that the potential is greatest in regions where there is the least need for additional food supply, and smallest in
regions with the greatest malnutrition challenges. For example, as of 2009, the study projects that the NAO region
and Europe could reduce FLW by 63 percent compared to Africa which could likely only reduce FLW by 31
percent (Kummu et al., 2012).
6.3 Characterization of FLW
This section looks at key similarities and differences in the FLW of the United States and other countries,
examining two factors that greatly affect the environmental footprint of FLW - when food is lost or wasted (i.e., at
what stage of the supply chain) and what categories of food are lost or wasted.
Supply Chain Stage
In general, lower-income nations lose a greater share of food during primary production, handling, and storage
than higher income nations, often due to insufficient infrastructure (e.g., cold chains) and technologies.
Conversely, higher-income nations, where consumers have more financial resources to purchase excess food,
waste a greater share of food during the consumption stage than lower income nations (FAO, 2019b; Spang et
al., 2019; FAO, 2013a, 2011). Note that this FLW during the consumption stage may be driven by forces beyond
individual and interpersonal factors, such as policies, marketing, media, or actions of the food industry (NASEM,
2020).
Consumption
Retail
Distribution &
Processing
Primary
Production
North America
and Oceania
Europe
Industrialized
Asia
North Africa,
West and
Central Asia
Latin America
South and
Southeast Asia
Sub-Saharan
Africa
FIGURE 6-2. SHARE OF CALORIES LOST AND WASTED, BY SUPPLY CHAIN STAGE,
FOR EACH GLOBAL REGION
In higher income, more developed regions like North America and Oceania the largest share of food loss and waste (FLW) is generated
during the consumption stage (i.e., households and food service). In lower income, less developed regions like Sub-Saharan Africa, the
primary production and distribution and processing (including storage) stages contribute the largest share of FLW.
Data Source: Lipinski etal. (2013)
Chapters. U.S. Food Waste in Global Context
65
-------
The FAO (2011) estimates that in iow-income countries, approximately 40 percent of FLW occurs during
production and processing, whereas in medium- and high-income countries, approximately 40 percent of FLW
occurs during retail and consumption. This pattern is further illustrated by the WRi analysis (Lipinski et al., 2013)
of the regional FAO (2011) data. Figure 6-2 shows the distribution of FLW by stage of the food system for each
world region. The production, handling and storage stages contribute substantially to FLW for less developed
regions like sub-Saharan Africa (72 percent), whereas the consumption stage contributes the largest share of
FLW in more developed regions, such as the NAO Region (58 percent). This is consistent with the U.S.-specific
estimates presented in Section 3.5.
Food Category
Globally and in the NAO region, fruits and vegetables comprise the largest share of FLW, and fish and seafood
comprise the smallest share.57 One key difference between NAO and the global average is the waste of animal
products, including meat, and milk and eggs; both categories account for a greater share of FLW in the NAO
region than in any other region (FAO, 2013a). See Figure 6-3 for the relative FLW of each food type by region,
according to FAO data (FAO, 2013a). Chen et al. (2020) provides additional detail at the country level, estimating
that the U.S. wastes 7.5 times more dairy, 3.5 times more meat, and 2 times more fruits and vegetables than the
global average.
Fish & Seafood C |
Meat & Poultry
Dairy & Eggs
Fruits
Vegetables
Grains
North America
and Oceania
Europe
Industrialized
Asia
North Africa,
West and
Central Asia
Latin America
South and
Southeast
Asia
Sub-Saharan
Africa
FIGURE 6-3. SHARE OF FLW, BY FOOD CATEGORY, FOR EACH GLOBAL REGION
Fruits and vegetables make up the largest share of food loss and waste (FLW) in every region. The loss and waste of animal products
(including meat, fish and seafood, milk and eggs) varies across regions, from 28 percent in North America and Oceania to 7 percent in
Sub-Saharan Africa.
Data Source: (FAO, 2013a, b)
57 Seafood losses are typically undercounted (Love et al., 2015).
Chapter 6. U.S. Food Waste in Global Context
66
-------
6.4 Per Person FLW
Per person measures allow for meaningful comparisons among countries of different sizes, thus the UN
Sustainable Development Goal (SDG) Target 12.3 for FLW and many national FLWgoals, including the U.S.
goal, are set on this basis. Three studies in the literature (Chen et al., 2020; Hig et al., 2016; Lipinski et al., 2013)
allow for comparison of global and U.S. (or NAO regional) per person edible FLW (i.e., each study produces a
global estimate and a U.S. or NAO estimate). While all three studies relied on FAO food availability data, their
scope and methodology of each study differs. Chen et al. (2020) examines consumption stage FLW exclusively,
while the other two studies include FLW from additional stages of the supply chain. See Figure 6-4 for a
comparison of the results. All studies indicate per person FLW in the U.S. is more than double the global average,
whether measured by weight or calories.
FLW By Calories
thousand calories / person / year
383
186
Global US
Chen et al. (2020)
Global US
Hig et al. (2016)
Global NAO
Lipinski et al. (2013)
Ei&l
FLW By Weight
kg / person / year
405
143
ipy
Itrl
Global US
Chen et al. (2020)
FIGURE 6-4. GLOBAL AND U.S. PER PERSON ANNUAL EDIBLE FLW
Food loss and waste (FLW) per person in the United States is more than double the global average, whether measured by weight or
calories. All these estimates include only edible FLW, excluding inedible parts such as bones or shells.
Hie et al. (2016) estimated per person FLW in calories in 111 countries in 2010 by calculating the difference
between food availability and the current dietary calorie requirements in each country. The authors calculated
calorie requirements for each country based upon demographic and anthropometric data (such as body weight)
from the UN and other sources.58 Hi? et al. demonstrated that the following eight countries exhibited food
surpluses, by calories, greater than 60 percent: the United States, plus Austria, Belgium, Egypt, France, Ireland,
Italy, and Turkey.
58 The study calculated country-specific dietary calorie requirements, noting the highest dietary requirements for the U.S.,
Lithuania, and United Arab Emirates of 2700-2800 calories/person/day.
Chapters. U.S. Food Waste in Global Context
67
-------
Lipinski et ai, (2013) highlights that per person edible FLW, by calories, in the NAO region (1,520 cal/person/day)
is double that of any other world region (other regions range from 414 to 748 cal/person/day). Chen et ai. (2020)
provides country-specific estimates of per person edible FLW59, by weight, finding that the United States
generates more than double the global average of per person edible FLW (503 g/person/day, as compared to 178
g/person/day). The authors also demonstrated that the U.S. ranked third highest of 151 countries examined in per
person edible FLW, behind Ireland and New Zealand.
Chen et al. (2020) went a step further to examine the nutrient composition (beyond calories) of the food that is lost
and wasted by each country during the consumption supplly chain stage (i.e., households and food service). The
authors estimated per person edible FLW for 151 countries in 2011 by combining edible amounts and nutrient
composition for 225 food items per country from the GENuS dataset (Smith et al. (2016), as cited in Chen et al.
(2020)) with region-specific FLW ratios from FAO. This analysis allowed them to produce a nutrition-related metric
for FLW - the "wasted daily diets" (WDD) contained in each country's per person FLW. A WDD represents the
number of daily nutritious diets (i.e., including recommended amounts of 25 nutrients) that could be provided
based on a country's per person FLW. By definition, this amount is equal to or lower than the number of daily
diets that could be provided based solely upon calorie requirements. Note that these estimates include only
consumption stage FLW60, thus underestimating the full potential of FLW.
Wasted Daily Diets (WDD)
1.00
FIGURE 6-5. WASTED DAILY DIETS
This figure shows a heatmap of Wasted Daily Diets (WDDs). Orange indicates the largest number of WDDs and blue indicates the fewest
number of WDDs. WDDs represent the number of daily nutritious diets (including recommended amount of 25 nutrients) that could be
provided based on a country's annual per person food waste at the consumption supply chain stage (i.e., households and food service).
Average global per person consumption-stage food waste embodies 18 WDDs; the United States embodies 41 WDDs, ranking fourth
highest among 151 countries.
Data Source: Chen et al. (2020)
59 Lipinski et al. (2013) and Chen et al. (2020) both used loss rates from Gustavsson et al. (2013); however, Lipinski et al.
(2013) used the 2009 FAO FBS data and Chen used the 2011 GENuS data set which is a disaggregated version of the 2009
FAO FBS data.
60 In the U.S., the consumption stage accounts for approximately one half of total FLW (ReFED, 2021a; Pagani et al.. 2020;
Vittuari et al., 2020; CEC, 2017).
Chapters. U.S. Food Waste in Global Context
68
-------
According to Chen at al. (2020)'s calculations, average global annual per person FLW at the consumption stage
embodies 18 WDDs, meaning an average person's consumption stage FLW over one year could fulfill the dietary
requirements for one person for 18 days (or 18 people for one day). In contrast, the U.S. per person FLW
embodies 41 WDDs,61 ranking fourth highest among 151 countries. Figure 6-5 shows a heatmap of the WDDs for
each country. The map shows that high-income countries, including the U.S., Canada, Australia, New Zealand,
and many of the European member states have the highest WDDs globally.
In general, affluence is highly correlated with higher per person FLW rates, regardless of the metric (e.g., weight,
calories or nutrients). The United States and other wealthy countries (i.e., upper-middle- and high-income
countries) all have higher per person FLW than less wealthy countries and the global average (Chen et al., 2020;
Xue and Liu, 2019; Vilarino et al., 2017; FAO, 2013a, 2011). Chen et al. (2020) demonstrated that mean FLW, by
weight, in high-income countries is almost two times that in upper-middle-income countries and four to six times
that in low-middle income and low-income countries. This pattern is consistent with a study by van den Bos
Verma et al. (2020) which found a logarithmic relationship between consumer affluence and FLW at the
consumption stage; at a threshold of affluence, the rate of consumption stage FLW increases with the level of
affluence. Hi
-------
6.5 Progress Towards Reducing Per Person FLW
As described in Chapter 1, the United States is one of many countries to adopt a national goal to halve per person
FLW at the retail and consumption stages by 2030, similar to the UN Sustainable Development Goal Target 12.3
(FAO, 2020). While the U.S. has not made progress toward halving FLW by 2030 (see Section 3.4), several
countries have reported significant reductions in FLW. For example:
• The United Kingdom reduced per person edible FLW by 27 percent, and total per person FLW (i.e.,
FLW with inedible parts included) by 21 percent between 2007 and 2018 (WRAP, 2020).
• The Netherlands reduced per person edible household FLW by 29 percent between 2010 and 2019
(Champions 12.3, 2017b).
• Norway reduced per person FLW across industry, wholesale, retail, and households by 12 percent
between 2010 and 2015, including an 11 percent reduction in per person edible household FLW
(Stensgard and Hanssen, 2016).
• Denmark reduced per person edible household FLW by 8 percent per person, and five percent in
total, between 2011 and 2017 (Danish EPA, 2018).
• Japan reduced household FLW by 13 percent between 2005 to 2009, achieving the majority of the
reduction in the first year (Parry et al., 2015).
These countries have taken a variety of actions to achieve these reductions, including setting goals, developing
national strategies with milestones, sponsoring educational campaigns, and developing partnerships with
organizations and businesses across the food system.
Additional data will likely become available as more countries establish FLW baselines and implement FLW
reduction programs. As of this writing, no country had announced that it had achieved Target 12.3, but the U.N.
Environment Programme has established the Food Waste Index to compare countries' progress toward the goal
(UN, 2021).
6.6 Global Environmental Footprint of FLW
The global food supply chain is a major driver of environmental degradation and natural resouce depletion.
Globally the food system uses 70% of all freshwater withdrawals, occupies about 40% of ice-free land, produces
34% of anthropogenic greenhouse gas (GHG) emissions, and is the largest contributor to biodiversity loss and
water pollution related to disruptions in the nitrogen and phosphorus cycles (Crippa et al., 2021; Tilman et al.,
2017; Diaz and Rosenberg, 2008; Ramankutty et al., 2008; Molden, 2007).
Global FLW places a tremendous burden on the planet, embedding roughly one quarter of the total global use of
cropland, freshwater resources, and fertilizers for food production, without accounting for the loss and waste of
animal products (Kummu et al., 2012). Global FLW also contributes 3.7 gigatons of CChe63 annually, not including
emissions related to landfills (Mbow et al., 2019; WRAP, 2015; FAO, 2013a). If FLW were a country, in 2010 it
would have been the third largest GHG emitter globally after China (21 percent of global emissions) and the
United States (13 percent of global emissions) (Ritchie and Roser, 2020; FAO, 2013a).
As world population and incomes rise, leading to dietary shifts, pressures on the environment from the food
system will rise as well. Studies project that food demand will increase by more than 50 percent between 2010
and 2050, and demand for more resource-intensive foods (i.e., animal products) will grow by nearly 70 percent
during the same time period (UN, 2020a; Searchinger et al., 2019). Reducing FLW provides one pathway toward
a more sustainable agricultural system (see Section 6.7).
As the U.S. chooses strategies to reduce FLW—and the environmental impact of FLW—it can be helpful for
policymakers to understand how the inputs and environmental impacts associated with U.S. FLW compare to
those of other countries.
63 With emissions associated with the disposal of FLW included, FLW accounts for 4.4 gigatons of CChe (FAO, 2015a, 2013a).
Chapter 6. U.S. Food Waste in Global Context
70
-------
Total Environmental Footprint of FLW
There are major differences in the magnitude of the environmental impact of FLW among countries, resulting from
differences in absolute and per person amounts of FLW (see Sections 6.1 and 6.4) as well as the breakdown
among the food categories lost or wasted and the supply chain stages at which food was lost or wasted in each
country (see Section 6.3). Environmental footprint increases as FLW increases, even more so if more of the FLW
happens downstream where embodied environmental impacts are greatest or more of the FLW is animal products
or fruits and vegetables (see Section 3.5 and Section 3.6 for discussion).
Given that the U.S. has greater FLW and greater per person FLW than the global average, and that downstream
FLW and animal product FLW comprise a greater share of U.S. FLW than of global FLW, it can be expected that
the environmental footprint of U.S. FLW will be substantially larger than the global average. Three studies (Chen
et al., 2020; Guo et al., 2020; Hi2 GHG emissions, blue
water use, cropland use, and nitrogen and phosphorous fertilizer application) of FLW per person per day in 2011
using global average characterization factors (i.e., the amount of environmental impact per gram of food) for 28
food groups. This method takes into account the amount and types of FLW in each country but does not account
for regional or country-specific environmental impacts (e.g., differences in climate or production methods).
Figure 6-7 and Figure 6-8 show detailed results for countries grouped by world region and income groups,
respectively, in order from highest to lowest per person daily FLW (Chen et al., 2020). Both figures also include
U.S.-specific estimates (calculated by the study's authors) for comparison purposes. As shown in Figure 6-7 and
Figure 6-8, high- and upper-middle-income countries and North America64, Europe and Central Asia, and East
Asia and the Pacific all had higher-than-world-average environmental footprint from producing wasted food (Chen
et al., 2020). Of note, the authors also found that GHG emissions per person from only the meat that was lost or
wasted in high-income countries exceeded the average GHG emissions per person of all FLW globally.
Chen et al. (2020) estimates that on average, global per person daily FLW is responsible for 124 g C02e (from
non-C02 GHG emissions), 58 L blue water, 0.36 m2 cropland, and 3.38 g fertilizer (combined phosphorous and
nitrogen). For comparison, per person daily FLW in the U.S. accounts for 457 g C02e, 151 L blue water, 103 m2
cropland, and 9 g fertilizer.
Examining the data at the country level, Chen et al. (2020) finds that per person U.S. FLW accounts for more blue
water use and fertilizer application than that of any of the other 150 countries evaluated. Additional studies report
that the U.S., along with China and India, accounts for the largest volumes of blue water associated with FLW
(Conrad et al., 2018; Birney et al., 2017; Lundqvist et al., 2008).
Per person U.S. FLW also accounts for the third highest amount of non-C02 GHG emissions and fourth highest
amount of cropland use, among all 151 countries. In addition, for each environmental criteria, the U.S. estimates
exceeded the per person average for high-income countries by more than 24 percent (Chen et al., 2020).
64 Chen et al. (2020) defines North America as U.S. and Canada.
Chapter 6. U.S. Food Waste in Global Context ^¦
71
-------
%T
P
a
Pr
FLW
Cropland
Freshwater
Fertilizer
Non-C02 GHG
Emissions
g
m2
L
g N + g P
g C02e
U.S.
503
103
151
9
457
North America
501
103
149
9
450
Europe & Central Asia
337
76
106
6
254
East Asia and Pacific
222
44
79
5
130
Middle East & North Africa
141
29
51
3
96
Latin America & Caribbean
121
28
45
3
143
South Asia
38
7
15
1
22
Sub-Saharan Africa
25
4
5
<1
17
FIGURE 6-7. DAILY PER PERSON FLW AND ASSOCIATED ENVIRONMENTAL FOOTPRINT,
BY GLOBAL REGION
Per person food waste during the consumption supply chain stage (i.e., households and food service) in the United States and its
associated cropland, freshwater, and fertilizer use, and greenhouse gas emissions exceed that of all world regions.
Data Source: Chen et al. (2020)
U.S.
High Income
Upper-Middle Income
Lower-Middle Income
Low Income
FLW
g
%T
Cropland
m2
P
Freshwater
L
Non-CO, GHG
Fertilizer _ . r
Emissions
g N + g P g C02e
503
103
151
9 457
FIGURE 6-8. DAILY PER PERSON FLW AND ASSOCIATED ENVIRONMENTAL FOOTPRINT,
BY GLOBAL INCOME GROUP
Food waste during the consumption supply chain stage (i.e., households and food service) in the United States and its associated
cropland, freshwater, and fertilizer use, and greenhouse gas emissions exceed the average for high-income countries.
Data Source: Chen et al. (2020)
Chapters. U.S. Food Waste in Global Context ^m 72
-------
Greenhouse Gas Emissions
Two other studies (Guo et al., 2020; Hi2e per person and 42.7 million MTCC^e total annually, similar to Western Europe (332 g CC>2e per
person) and less than Australia & New Zealand, South America, and Northern Europe (848, 684, and 407 g CC>2e
per person, respectively). Northern America also ranked fourth for total emissions, behind Eastern Asia, South
America, and Southern Asia, accounting for approximately 8 percent of global non-CC>2 GHG emissions from
FLW. The estimates of Hi
-------
More recently, Guo et al. (2020) determined the region- and country-specific GHGs associated with FLW using
updated FAO FLW data for 2017 (FAO, 2019a). Guo et al. (2020) used an expanded scope, including emissions
of all GHGs from primary production and international transportation - as compared to Chen et al. (2020) and Hi2e per year) in the world after China and India. Note that the United States is the only developed
country67 on the list of top 10 FLW-GHG-generating countries, indicating that all other developed countries have
lower rates of FLW-associated GHG emissions generation. The top 10 countries (China, India, United States,
Indonesia, Brazil, Nigeria, Russia, Pakistan, Mexico and Malaysia) account for approximately 60 percent of global
FLW and FLW-associated GHG emissions (Guo et al., 2020).
6.7 Environmental Benefits of Halving Global FLW
The potential environmental benefits of halving global FLW are similar to those of halving FLW domestically (see
Chapter 5) - decreasing key inputs and environmental impacts such as the use of land for agriculture; use of blue
water, fertilizer, and energy; and GHG emissions. Researchers also agree that reducing FLW will increase the
production and consumption efficiency of the food system (i.e., increase the amount of food produced per unit of
resources) (Cattaneo et al., 2020; Pagani et al., 2020).68 This increased efficiency from FLW reductions can
increase the amount of food that can be produced for the same impacts, which will be increasingly important, as
the world population and incomes grow, leading to increased global demand for food, and particularly demand for
foods with greater environmental impacts such as meat, dairy and processed foods (Searchinger et al., 2019;
Tilman and Clark, 2014; Godfray et al., 2010). Reducing FLW can help meet increased global demand for food
without the full projected increase in environmental impacts. It is not expected that the current rate of yield
increases can meet expected food demand in 2050 without further deforestation and biodiversity loss (Ray et al.,
2013; Garnett, 2011).
Over the past two decades, eight studies have estimated the environmental benefits of reducing FLW globally.
While all of the studies used the FAO (2013a, 2011) food balance sheets and FLW rates, the studies vary in
geographic coverage, years covered, and methods to estimate the environmental impacts of producing food and
reducing FLW. Note that all the studies estimate maximum environmental benefits and do not consider economic
factors that may impede the realization of benefits.
A key difference among the studies' methodologies is that the first two studies (Jalava et al., 2016; Kummu et al.,
2012) estimate savings associated with reducing FLW relative to the current environmental footprint of the food
system, while the latter six studies (Clark et al., 2020 ; Searchinger et al., 2019; Springmann et al., 2018; Roos et
al., 2017; Bajzelj et al., 2014; Tilman and Clark, 2014) estimate savings relative to a projected future BAU
scenario. The results of all studies are presented in Table 6-1.
67 As classified by the United Nations.
68 Using wheat as a hypothetical example, producing food (e.g., growing 10 acres of wheat) requires inputs (e.g., water, land,
energy, crop inputs) and results in environmental impacts (e.g., GHG emissions, water quality impacts). At a current FLW rate
of 30 percent, out of 30,000 pounds of wheat produced on 10 acres, only 21,000 pounds are currently consumed, meaning the
current efficiency is: (21,000 pounds wheat/[environmental footprint]). If FLW rates were reduced by half overall (e.g., 15
percent FLW) the efficiency of wheat production would increase: (25,500 pounds wheat/[environmental footprint]). Assuming
that the reduction in FLW does not result in increased resource use and change the environmental footprint.
Chapter 6. U.S. Food Waste in Global Context
74
-------
TABLE 6-1. MAXIMUM ENVIRONMENTAL BENEFITS OF HALVING GLOBAL FOOD WASTE
Environmental Savings
Relative to Current _ . .. . oncncAno 2020 Relative
0 . . . Relative to 2050 BAU Scenario . 0„nn
Food System Footprint to 2100
Environmental
Impact
Land a
(million m2)
Waterb
(million L)
Fertilizer
(million kg)
Jalava
et al. (2016)
Kummu et al.
(2012)
Bajzelj
etal. (2014)
Roos
et al. (2017)
Searchinger
etal. (2019)
Springmann
etal. (2018)
Tilman & Clark
(2014)
Clark
et al. (2020)d
766,410
6,400,000
3,140,000
2,891,000
—
77,940,000
960,000,000
—
—
387,702,000
—
—
11,691
18,000
—
—
29,280 c
—
—
— —
4,500
1,500
614
6%
500
4,134*
GHG
(million MTCChe)
All studies examined food loss and waste from all four cradle-to-consumer supply chain stages. However, Kummu et al. (2012) excluded animal products (and the pastureland and feed needed to produce them).
aln all studies, except for Kummu et al. (2012) and Springmann etal. (2018), total land reduction is based on the combined average land reduction of cropland and pasture. Kummu etal. (2012) and Springmann et
al. (2018) only included cropland.
b Water only includes "blue water" (water used for irrigation). All estimates are for consumptive use except for Bajzelji et al. (2014) which used data on water withdrawals.
c Values are 17% reduction for nitrogen fertilizer and 16% reduction for phosphorus fertilizer. Potassium fertilizer is not included in the analysis.
dThe study also reports an estimate of 27% reduction as CO2 warming-equivalent (C02-we). The Clark etal. (2020) cumulative savings value of 331 cumulative GWP100 Gt C02e was divided by the time period (80
years) to calculate an annualized estimate.
Chapter 5. Benefits of Reducing U.S. Food Waste
75
-------
Environmental Benefits, Relative to the Current Footprint
Two studies (Jalava et al., 2016; Kummu et al., 2012) estimate the potential environmental benefits of halving
global FLW, relative to the current environmental footprint of the cradle-to-consumer food system. Kummu et al.
(2012) estimated the maximum cropland (excluding feed crops), blue water, and fertilizer savings that could be
achieved in a global "minimum possible FLW' scenario. The study defines this scenario as each of the seven
world regions achieving the lowest current FLW rate (among all regions) for each food category in each supply
chain stage, as described in Section 5.1 Under this scenario, global FLW is reduced by 48 percent. Kummu et al.
(2012), however, excludes the loss and waste of animal products, which have significant environmental footprints
relative to other food categories (see Chapter 2 and Chapter 4).
Kummu et al. (2012) estimated that, globally, the amount of water currently wasted due to FLW could be reduced
by 44 percent, wasted cropland could be reduced by 39 percent, and wasted fertilizer could be reduced by 42
percent. For the NAO region specifically, the study found that greater reductions were possible, estimating that
the amount of water currently wasted due to FLW could be reduced by 57 percent, wasted cropland could be
reduced by 53 percent, and wasted fertilizer could be reduced by 54 percent. The ability of the NAO region to
make greater reductions than the global average is driven by the region's ability to reduce the amount of food lost
or wasted by 63 percent, as compared to a global average reduction of 48 percent (i.e., it is due to its higher
current FLW rate). Note that these values are the percent reduction of wasted inputs (i.e., not total inputs) that
would occur from reducing FLW. The values do not show the percent reduction that would occur compared to the
whole environmental footprint of the global food supply system.
Table 6-1 compares Kummu et al. (2012) results with the footprint of the global food supply system, showing an
11 percent reduction of water use, a 9 percent reduction of land use, and a 10 percent reduction of fertilizer use
from reducing FLW globally. In comparison, Kummu et al. (2012) estimates the NAO region can reduce the
environmental footprint of its food system by 14 to 15 percent by reducing FLW (see Section 5.1 for discussion).
Jalava et al. (2016) looked exclusively at potential water use savings from halving FLW. The study applied water
use data from the Water Footprint Network to estimate the impact of halving FLW on water use in each of the
same seven global regions considered by Kummu et al. (2012). The authors estimated an overall global reduction
of 12 percent of blue water use, ranging from an 11 percent reduction in South and Southeast Asia to a 15
percent reduction in the Middle East and North African region and Latin America. The study also estimated a 12
percent global reduction of green water use. The authors also provided country-level estimates. For the U.S.
specifically, Jalava et al. (2016) estimated one percent greater blue water reduction potential (i.e., 13 percent
compared to 12 percent) and equal potential for green water reductions (i.e., 12 percent) as compared to the
global average.
Environmental Benefits, Relative to the Future Footprint
As the human population continues to grow, there is concern that the increases in food production required to
feed the growing population may exceed global environmental limits. In addition, as incomes rise, global food
production is projected to shift to a higher percentage of more resource-intensive foods, such as animal products
and fruits and vegetables, leading to a larger environmental footprint than the current mix of food produced
(Springmann et al., 2018; Tilman and Clark, 2014).
As such, six studies (Clark et al., 2020; Searchinger et al., 2019; Springmann et al., 2018; Roos et al., 2017;
Bajzelj et al., 2014; Tilman and Clark, 2014) estimated the future environmental impacts of the global food system
by comparing a BAU scenario to a scenario in which FLW is halved. FLW data were derived from FAO food
balance sheets in all studies. The BAU scenarios were based upon UN midrange (i.e., medium fertility) population
estimates and demonstrate increasing environmental pressures in the absence of new mitigation strategies, such
as major technological advances, dietary shifts, or FLW reductions. Note that while the studies include projections
of economic and consumption trends, they do not account for economic factors such as rebound effects. All the
studies predicted that all studied aspects of the environmental footprint of the global food system would increase
from 2009 to 2050 under the BAU and halving FLW scenarios, but that halving FLW would decrease the impact
compared with the BAU scenario.
Chapter 6. U.S. Food Waste in Global Context
76
-------
Five of the studies examined reductions from 2050 BAU scenario (Searchinger et al., 2019;Springmann et al.,
2018; Roos et al., 2017; Bajzelj et al., 2014; Tilman and Clark, 2014). The studies used a wide array of
assumptions and methods to develop their 2050 BAU scenario. To estimate demand, the five studies used similar
estimates of population growth but varying methods and data sources for socio-economic changes and
accompanying changes in dietary patterns. To estimate production and its environmental impacts, some studies
(Springmann et al., 2018; Roos et al., 2017) kept current conditions while others (Searchinger et al., 2019; Bajzelj
et al., 2014; Tilman and Clark, 2014) extended current trajectories (to capture expected improvements) in
agricultural efficiencies, such as yield increases and nitrogen use efficiency gains.
As shown in Table 6-1, the studies estimate halving FLW could reduce projections of agricultural land
requirements, water use, and fertilizer use, and greenhouse gas emissions by 6 to 22 percent, depending on the
study's boundary. Expected savings in water (12 to 15 percent) and fertilizer (12 to 16 percent) use were fairly
consistent across studies; however estimates of agricultural land use and greenhouse gas emissions varied more
widely, likely due to differences in study scopes.
When measuring changes in land use, one study (Springmann et al., 2018) looked exclusively at cropland while
others (Bajzelj et al., 2014; Searchinger et al., 2019;Roos et al., 2017) included pasture land as well, dramatically
increasing the amount of land being evaluated. Differences in absolute land savings in Bajzelji et al. (2014) and
Searchinger et al. (2019) are most likely attributable to differences in methods for modeling future land use shifts
and productivity gains. Searchinger et al. (2019) projected increased cropping intensities, more efficienct grazing,
and improved pasture productivity, which results in greater livestock output per hectare. Additionally, Searchinger
et al. (2019) linked expected productivity gains in BAU 2050 with limitations on shifting agricultural land into new
regions.
The study by Bajzelj et al. (2014) is unique in that it also estimated the changes in net forest cover and tropical
pristine forests associated with halving FLW, as shown in Figure 6-10. To measure this change, Bajzelj et al.
(2014) considered the distribution of land expansion across different biomes and used data on estimated
agricultural land expansion, current land use, and agricultural suitability of land. Overlaying this data allowed for
predictions in forest losses.
The studies also differed in how they addressed greenhouse gas emissions. While Springmann et al. (2018)
looked exclusively at ChU and N2O emissions, projecting a 6 percent decrease; the other four studies considered
all greenhouse gas emissions and found larger reductions possible (9 to 22 percent). The studies also varied in
their boundaries for which activities were covered. All studies considered emissions from enteric fermentation and
rice cultivation, for example, but only two of the five studies (Searchinger et al., 2019; Bajzelj et al., 2014) also
included emissions from land use change. Emissions from land use change can vary considerably based on
where land expansion occurs, and the two studies considering emissions from land use change maped land
expansion differently. Additionally, Searchinger et al. (2019) assumed that in the BAU 2050 scenario reforestation
of lands with little agricultural potential would provide offsets for agricultural land expansion, potentially leading to
lower reductions compared to Bajzelji et al. (2014). Studies also differed in their expected emissions factors for
certain activities. The study by Roos et al. (2017) was unique in that they included a 10% reduction in emissions
from livestock associated with expected breeding and feeding improvements realized by 2050. Searchinger et al.
(2019) included a 25% decrease in emissions from on-farm energy use, projecting a shift away from fossil energy
sources. Given the myriad assumptions used, future projections remain largely uncertain, yet all studies
consistently showed that halving food loss and waste was an integral intervention for reducing food system
impacts.
Considering the U.S. specifically, Springmann et al. (2018) estimated roughly similar reductions (i.e., within one
percent) for the U.S. and the world for land, water, and fertilizer use; however, the study estimated the United
States could achieve a 2 percent greater reduction in CH4 and N2O emissions from primary production, relative to
BAU in 2050, likely due to the greater amount of animal products wasted in the U.S. The other studies did not
provide country-specific projections.
Chapter 6. U.S. Food Waste in Global Context
77
-------
50%
Tropical
Pristine
Forests
Net Forest
Cover
Cropland
Pasture
-20%
Current Trajectory in 2050 Halving Food Waste
FIGURE 6-10. PROJECTED BENEFITS FROM HALVING GLOBAL FLW
To meet rising food demands associated with population growth and changing dietary patterns, agricultural land will need to expand in
2050 (+42 percent cropland, +13 percent pasture). Increases in land for agriculture will require reductions in total net forest coverage (-14
percent) and tropical pristine forest (-10 percent). If global food loss and waste were halved, agricultural land expansion could be
attenuated by 19 percent for cropland and 10 percent for pasture, and encroachment on net forest cover and tropical pristine forests could
be reduced by 5 percent and 1 percent, respectively. Data Source: Bajzelj et al. (2014)
Clark et al. (2020) covers a different time period than the other two studies, estimating cumulative reductions
between 2020 and 2100. The authors focused exclusively on GHG emissions reductions during primary
production (including emissions from land use change), finding that gradually halving FLW between 2020 and
2050 could result in a 24 percent cumulative reduction in primary production GHG emissions, when measured in
100-year global warming potential, and 27 percent cumulative reduction, when measured as warming-equivalents
(CC>2-we), compared with the BAU scenario (Clark et al., 2020). This study primarily reported "warming
equivalents," rather than the 100-year global warming potential like all other studies presented in this report, to
account for the short-lived nature of methane. Between 2020 and 2100 the authors project savings of 331 Gt
C02e (364 Gt C02-we) during primary production from halving food loss and waste (Clark et al., 2020).
Importantly, Clark et al. (2020) analysis shows that food system emissions reductions are essential to
achievement of limiting global warming to below 1.5° and 2° Celsius, compared to pre-industrial levels. Even if
fossil fuel emissions were immediately halted, current trends in the food system (including increasing yields at the
current rate) could preclude the achievement of these targets. The analysis shows this FLW-related reduction in
GHGs increases the likelihood that the global climate will remain below the maximum increase of 2°C target
outlined in the Paris Agreement (UNFCC, 2015).
Many of these studies compared a variety of strategies—including closing yield gaps, increasing resource
efficiency, dietary shifts, and reducing FLW—finding that only in combinations could these strategies achieve a
sustainable agricultural future (Clark et al., 2020; Searchinger et al., 2019; Springmann et al., 2018; Roos et al.,
2017; Bajzelj et al., 2014; Tilman and Clark, 2014).
Chapter 6. U.S. Food Waste in Global Context
78
-------
CHAPTER 7.
Conclusions and Research Gaps
This report summarizes available data on the cradle-to-consumer (farm-to-kitchen) environmental footprint of U.S.
FLW and the potential environmental benefits that could be realized by reducing FLW. General conclusions are
presented in Section 7.1, while specific strategies to maximize the environmental benefits of FLW reduction
efforts are presented in Section 7.2. The report concludes with identification of priority research areas in Section
7.3.
7.1 Conclusions
More than one-third of the U.S. food supply is not consumed, resulting in a "waste" of resources—including
agricultural land, water, pesticides, fertilizers, and energy—and the generation of environmental impacts—
including greenhouse gas emissions and climate change, consumption and degradation of freshwater resources,
loss of biodiversity and ecosystem services, and degradation of soil quality and air quality (U.S. EPA, 2019b; IOM
and NRC, 2015; U.S. EPA, 2015).
Despite differences in study design, methodologies, data sets, time periods, and other factors, most estimates of
the environmental impacts associated with food that is ultimately lost or wasted show general agreement once
these factors are taken into consideration. The most comprehensive credible estimates available in the literature
estimate that each year, uneaten food in the United States embodies:
• 560,000 km2 (140 million acres) agricultural land - approximately 16 percent of U.S. agricultural land
(Read et al„ 2020);
• 22 trillion L (5.9 trillion gallons) blue water - equal to the annual water use of more than 50 million
American homes (Read et al., 2020);
• 350 million kg (778 million pounds) pesticides (Conrad et al., 2018);
• 6,350 million kg (14 billion pounds) fertilizer (Toth and Dou, 2016);
• 2,400 million GJ (664 billion kWh) energy - enough energy to power more than 50 million U.S. homes
(Pagani, et al., 2020; Vittuari et al., 2020); and
• 170 million MTCChe GHG emissions (excluding landfill emissions) each year - equal to emissions of 42
coal-fired power plants (Read et al., 2020).
This uneaten food also contains enough calories to feed more than 150 million people each year, far more than
the 35 million estimated Americans experiencing food insecurity (USDA, 2021a; Wood et al., 2019; Buzby et al.,
2014). While the estimates presented above were chosen largely based upon their comprehensive scope, many
credible methodologies presented in this report would produce higher estimates if their results were extrapolated
to cover FLW along the whole supply chain. The estimates above from Read et al. (2020) in particular should be
considered conservative.
As global populations and incomes rise, and the environment faces pressures from increased food production,
reducing the per person environmental footprint of agriculture will be essential to the sustainability of the planet
(Clark et al., 2020; Springmann et al., 2018). Limited options are available to sustainably increase the global food
supply to meet growing demand. Closing yield gaps and increasing productivity alone will likely be insufficient to
prevent further deforestation and environmental degradation (Hayek et al., 2021; Bajzelj et al., 2014). Even under
the most promising scenarios of yield increases, up to 20 percent more land will be needed by 2050 (Bajzelj et al.,
2014). Thus demand-side measures, such as reducing FLW or dietary shifts, will also be needed to sustainably
increase the food supply (Rosenzweig et al., 2021; Wood et al., 2019; Roos et al., 2017; Foley et al., 2011). Many
researchers have noted that policymakers may find reducing FLW less controversial and more tractable than
dietary shifts (Birney et al., 2017; Neff et al., 2015; Smith et al., 2013).
Chapter 7. Conclusions and Research Gaps
79
-------
While halving FLW cannot alone make the global food system sustainable, it can play a significant role and may
well be essential (Gerten et al., 2020). The most comprehensive (though likely conservative, as noted above)
estimates of annual environmental savings (i.e., avoided inputs and environmental impacts) from halving U.S.
FLW include:
• More than 300,000 square km2 (75 million acres) agricultural land - an area greater than the State of
Arizona;
• 12 trillion L (3.2 trillion gallons) blue water - equal to the water use of 29 million American homes;
• Nearly 290,000 metric tons (640 million pounds) of bioavailable nitrogen from agricultural fertilizer with
the potential to reach a body of water, cause algal blooms and deteriorate water quality;
• 940 million GJ (262 trillion kWh) energy - enough to power 21.5 million U.S. homes; and
• 92 million MTCC^e GHG - equal to the CO2 emissions from 23 coal fired power plants (Read et al.,
2020).
These savings can only be achieved through prevention (i.e., source reduction) of FLW. Recycling of food waste
cannot achieve these benefits since a substantial fraction of the impacts occur during the primary production of
food.
In global context, the United States is a major producer of food loss and waste, wasting more food (total) and
more food per person than most other nations. Only two countries generate more food waste and more food
waste per person than the United States (China and India, and New Zealand and Ireland, respectively) (Guo et al,
2020; Chen et al., 2020). The environmental impact of U.S. food loss and waste is also substantial relative to
other countries, as the U.S. wastes more food downstream and more animal products than the global average
(Lipinski et al., 2013; FAO, 2013a, b). Studies consistently show the United States using more resources and
creating more environmental impacts per unit of food waste than the global average (Chen et al., 2020; Guo et al,
2020; Hi? et al., 2016).
The United States is one of many countries to adopt a national goal to halve per person FLW at the retail and
consumption stages by 2030, similar to the UN Sustainable Development Goal Target 12.3 (FAO, 2020). While
the U.S. has not made progress toward halving FLW by 2030, examples of significant progress in similar
countries are emerging. For example, the UK has reduced per person edible FLW by 27 percent, achieving the
bulk of the reduction within four years (WRAP, 2020), and Japan has reduced household FLW by 13 percent in
four years, achieving the majority of progress in one year (Parry et al., 2015).
Scientists project halving global FLW could result in a 24 percent reduction in cumulative global food system
greenhouse gas emissions between 2020 and 2100 (331 Gt C02e), relative to a business-as-usual scenario
(Clark et al., 2020). Significant reductions (6 to 16 percent) could also be achieved in the amounts of agricultural
land, water, and fertilizer used in 2050 (compared to business-as-usual scenario) by halving global food loss and
waste (Searchinger et al., 2019; Springmann et al., 2018; Roos et al., 2017; Jalava et al., 2016; Bajzelj et al.,
2014; Kummu et al., 2012).
Many of the studies presented in this report compared a variety of strategies—including closing yield gaps,
increasing resource efficiency, dietary shifts, and reducing FLW—finding that only in combinations could these
strategies achieve a sustainable agricultural future (Clark et al., 2020; Searchinger et al., 2019; Springmann et al.,
2018; Roos et al., 2017; Bajzelj et al., 2014; Tilman and Clark, 2014). This report demonstrates the substantial
contribution halving food loss and waste, both domestically and internationally.
Chapter 7. Conclusions and Research Gaps
80
-------
7.2 Strategies to Maximize the Environmental Benefits of
Halving U.S. FLW
The recent literature reviewed in this report provides many insights as to how policymakers might tailor FLW
programs and policies to maximize environmental benefits. Specifically, it provides three key guiding principles:
(1) The greatest environmental benefits can be achieved through prevention rather than
recycling. Given that significant inputs and environmental impacts (use of land, water, pesticides and
fertilizer, plus GHG emissions) associated with FLW occur during primary production, the greatest
benefits can be achieved by avoiding the production of unnecessary food (or lessening the need for
additional food production). The estimates of the maximum environmental benefits of reducing FLW
presented in this report all rely upon the assumption that decreases in demand (from reducing FLW)
will result in equivalent decreases in production. While not all prevention activities may achieve this
(due to economic factors outside the scope of this report), wasting food and then recycling it does not
provide a similar opportunity to achieve these maximum benefits. Recycling will not signal demand for
a smaller per person food supply or "undo" the impacts of primary production.
(2) The largest energy and greenhouse gas emissions benefits can be obtained by reducing
consumption stage (households and restaurants) FLW. In the U.S. the consumption stage
represents roughly half of all FLW (ReFED, 2021 a; Pagani et al., 2020; Vittuari et al., 2020; CEC,
2017) and, as the last stage in the cradle-to-consumer supply chain, accounts for the greatest
environmental impacts (since impacts are cumulative). While use of land for agriculture and the use
of water, pesticides and fertilizers occur chiefly during primary production, energy use and GHG
emissions occur all along the supply chain and thus the embodied energy use and GHG emissions
increase as food moves along the supply chain. In a study projecting the environmental benefits of
halving U.S. FLW in each of seven sectors, the authors found that the bulk of the environmental
benefits could be achieved by halving FLW in only three sectors: households, restaurants, and food
processing (Read et al., 2020). The same study suggests that a focus on institutional food service
(e.g., schools or hospitals) or retail (also downstream sectors) will yield minimal environmental
results. Note, however, that upstream factors can drive consumption stage waste, and solutions to
reducing FLW in one sector may be implemented in that sector or upstream from that sector (e.g.,
reducing consumer FLW by making changes in supermarkets and restaurants) (NASEM, 2020).
(3) Focusing on reducing FLW of the most resource-intensive foods, such as animal products
and fruits and vegetables, can yield the greatest environmental benefits. These two categories
consistently rank as top contributors to many of the environmental impacts associated with FLW.
Animal products (especially beef) have a particularly outsized contribution. Making up less than one-
third of U.S. FLW, animal products are responsible for the largest share of agricultural land use,
nitrogen and phosphorous fertilizer application, energy use, and GHG emissions, plus one quarter of
pesticide application and at least a third of blue water use (Chen et al., 2020; Guo et al., 2020; Pagani
et al., 2020; Vittuari et al., 2020; Wood et al., 2019; Conrad et al., 2018; Mekonnen and Fulton, 2018;
Birney et al., 2017; Toth and Dou, 2016; Heller and Keoleian, 2015; Buzby et al., 2014; Gustavsson
et al., 2013; Cuellar and Webber, 2010). Fruits and vegetables make up a larger portion of U.S. FLW
than animal products and are the leading contributor to pesticide application and blue water use
associated with FLW (Conrad et al., 2018; Mekonnen and Fulton, 2018; Birney et al., 2017). Fruits
and vegetables are also responsible for the second largest share of fertilizer application, behind
animal products (Chen et al., 2020; Wood et al., 2019; Conrad et al., 2018). Thus, achieving
reductions in the loss and waste of these food categories should have more substantial
environmental benefits than in other food categories. Studies projecting environmental benefits of
halving FLW largely confirm this proposition, adding grains as an additional FLW food category worth
attention (Read et al., 2020) and noting the potential of reducing FLW of animal products to also most
significantly reduce ammonia emissions from nitrogen fertilizer application (Wood et al., 2019).
Chapter 7. Conclusions and Research Gaps
81
-------
7.3 Research Gaps
In recent years, a myriad of domestic and international food loss and waste reduction initiatives, including the U.S.
2030 Food Loss and Waste Reduction Goal, have spurred research on understanding and reducing FLW.
However, there are still data gaps and uncertainties associated with estimating the amount and characteristics of
FLW, the embodied environmental footprint of FLW, and the potential environmental benefits from FLW
reductions. Further research is needed to refine estimation methods and to improve the availability, quality,
consistency, and frequency of updates of necessary data. In addition, a deeper understanding of the interplay
between supply chain stages and the unique drivers of FLW in the United States could lead to more successful
policies and programs. Science-based answers to these research questions and others will increase U.S.
policymakers' understanding of U.S. FLW and help them to tailor FLW strategies to meeting the Food Loss and
Waste Reduction Goal with maximum environmental benefits.
EPA is currently undertaking projects to advance FLW knowledge in three areas relevant to this report, including:
Evaluating the comprehensive net environmental footprint of U.S. FLW. EPA will integrate
the data in this report (the cradle-to-consumer footprint) with data from EPA's forthcoming report
on the environmental impacts of U.S. food waste management pathways (such as landfill,
combustion, composting and anaerobic digestion) to assess the net environmental footprint of
U.S. FLW from cradle-to-grave.
Creating environmental indicators to track the environmental footprint of U.S. FLW over
time. EPA will develop indicators, as a part of the Report on the Environment, to track the amount
of FLW and its associated inputs and environmental impacts overtime, beginning with
greenhouse gas emissions.
Enhancing modeling of the food system including the generation of FLW by updating the
USEEIO model and creating a WARM/USEEIO hybrid model. EPA is partnering with USDA,
Argonne National Laboratory, and Cornell University to build a more refined food system model
within the USEEIO. Additionally, EPA is working to incorporate additional life cycle analysis data
on the end-of-life management pathways for food waste and their associated environmental
impacts.
New, original research is needed to fill these additional priority knowledge gaps:
Enhance the data on U.S. FLW and address data gaps. Research is needed to update data,
address data gaps, and understand key differences among FAO, USDA, and EPA estimates of
the amount, food category, and supply chain stage at which food is currently lost or wasted in the
United States. The accuracy of all the data sets could be improved. Additional data needs to be
collected to inform more precise estimates of FLW during primary production (including fisheries
and aquaculture) and food processing.
Increase frequency at which the United States can track progress in reducing FLW.
Methods and tools need to be developed to track changes in the amount of edible and inedible
FLW in the United States with greater frequency and without the use of static FLW rates.
Currently FLW rates (also known as loss factors) for each food category are updated once a
decade, making it difficult to assess progress and gauge success of programs to reduce FLW.
Quantify the environmental impacts occurring in other countries that are associated with
U.S. waste of imported foods. Research is needed to estimate the environmental impacts,
including deforestation (and associated carbon and biodiversity loss) and water scarcity, of
producing food that is exported to and ultimately wasted in the United States. Linking FLW to
specific locations of environmental degradation or resource use can help policymakers target
FLW programs to decrease pressure to convert land to agriculture, especially in the tropics, or to
use limited water resources in water-stressed areas. The estimates of inputs and environmental
impacts presented in this report generally assume the entire U.S. food supply is produced
domestically and do not account for differences in local environments, agricultural practices,
standards, and production methods among countries. This data would also help to refine
projections of environmental savings possible through halving U.S. FLW.
Chapter 7. Conclusions and Research Gaps
82
-------
Strengthen our understanding of the interaction among food system supply chain stages
with regard to FLW. Food is produced and supplied by a complex, multistage system that
includes many industries and participants. As a whole, the food system functions not as a
sequential farm-to-kitchen food chain but as a complex system of interdependencies and
feedbacks that responds to myriad economic, environmental, and social factors. Research is
needed to assess how changes in downstream demand affect upstream production and supply
(e.g., under what conditions will reducing consumer food demand decrease production/supply
accordingly, thus resulting in maximum environmental savings), including the effect of economic
factors. Research is also needed to identify where solutions to FLW in one stage may be best
implemented upstream from that stage (e.g., manufacturers and supermarkets making changes
to help households waste less food).
Explore how current trends in the U.S. food system will affect FLW and its environmental
footprint in the future. Research is needed to project the impact of trends such as the
increasing use of online grocery shopping and changes in household size, on the amount and
characteristics of U.S. FLW in order to help policymakers design successful long-term FLW
reduction strategies.
Evaluate the life cycle impacts of proposed FLW prevention strategies. To achieve the full
environmental benefits projected in this report, half of current U.S. FLW must be prevented, not
just recycled. A variety of FLW prevention solutions exist and new ideas and technologies are
coming online. In some cases, these solutions may require resources or present potential
environmental impacts. Research is needed to estimate the net benefits of solutions such as the
use of innovative or additional packaging or choosing frozen, canned, or dried produce over fresh
produce to inform decisions and policies to reduce FLW and its environmental impacts.
Improve precision of food loss and waste estimates. Additional data on loss and waste of
specific food types (within the food categories currently measured) would allow link to more
precise life cycle analysis data and improve the accuracy of environmental impact estimates as
well as allow for more specific targeting of higher impact foods during FLW interventions, such as
education about optimal storage conditions.
Deepen our understanding of drivers of FLW unique to the United States. While many of the
key drivers of FLW are common across similar countries, this report identified several factors
unique to the U.S. which impact FLW. For example, the U.S. food supply per person is far greater
than in other parts of the world. Identifying and examining systemic or institutional contributors to
U.S. FLW can lead to more successful solutions. Research is needed to explore drivers of over-
supply of food (i.e., supply beyond demand) in the United States, beyond levels seen in other
developed, high-income countries, and to identify unique forces in the U.S. food system that
consistently result in significant food loss and waste. Research could also lead to improved
management methods for surplus food.
Evaluate effectiveness of policy and program options to reduce FLW. Research and
evaluation are needed to determine which types of interventions could be the most successful in
reducing FLW—and its associated environmental impacts—and to identify synergies and conflicts
with other policy objectives, such as improving health and reducing food insecurity.
Chapter 7. Conclusions and Research Gaps
83
-------
References
Aktar, W; Sengupta, D; Chowdhury, A. (2009). Impact of pesticides use in agriculture: their benefits and hazards.
Interdisciplinary Toxicology 2: 1-12. https://doi.orq/10.2478/v10102-009-0001 -7
Aleksandrowicz, L; Green, R; Joy, EJ; Smith, P; Haines, A. (2016). The impacts of dietary change on greenhouse
gas emissions, land use, water use, and health: A systematic review. PLoS One 11: e0165797.
https://doi.org/10.1371/iournal.pone.0165797
Aneja, VP; Schlesinger, WH; Erisman, JW. (2009). Effects of Agriculture upon the Air Quality and Climate:
Research, Policy, and Regulations. Environmental Science & Technology 43: 4234-4240.
httPs://doi.ora/10.1021/es8024403
Asem-Hiablie, S; Battagliese, T; Stackhouse-Lawson, KR; Alan Rotz, C. (2019). A life cycle assessment of the
environmental impacts of a beef system in the USA. The International Journal of Life Cycle Assessment
24: 441-455. https://doi.ora/10.1007/s11367-018-1464-6
Atwood, D; Paisley-Jones, C. (2017). Pesticides industry sales and usage: 2008-2012 Market Estimates.
Washington, DC. https://www.epa.gov/sites/production/files/2017-01/documents/pesticides-industrv-
sales-usaqe-2016 O.pdf
Avadi Tapia, AD; Henriksson, PJG; Vazquez-Rowe, I. (2016). Challenges and best practices of seafood supply
chains LCA. LCA Food 2016: 10th International conference on Life Cycle Assessment of Food, 2016-10-
19/2016-10-21, Dublin, Ireland.
Bajzelj, B; Quested, TE; Roos, E; Swannell, RPJ. (2020). The role of reducing food waste for resilient food
systems. Ecosystem Services 45: 101140.
https://www.sciencedirect.com/science/article/pii/S221204162030Q826
Bajzelj, B; Richards, KS; Allwood, JM; Smith, P; Dennis, JS; Curmi, E; Gilligan, CA. (2014). Importance of food-
demand management for climate mitigation. Nature Climate Change 4: 924-929.
https://doi.ora/10.1038/nclimate2353
Bhagwat, V. (2019). Safety of Water Used in Food Production. In Food Safety and Human Health: Academic
Press. httPs://doi.oro/10.1016/B978-0-12-816333-7.00009-6
Birney, C; Franklin, K; Davidson, T; Webber, M. (2017). An assessment of individual foodprints attributed to diets
and food waste in the United States. Environmental Research Letters 12: 105008.
https://doi.ora/10.1088/1748-9326/aa8494
Boehm, R; Wilde, PE; Ver Ploeg, M; Costello, C; Cash, SB. (2018). A comprehensive life cycle assessment of
greenhouse gas emissions from U.S. household food choices. Food Policy 79: 67-76.
http://www.sciencedirect.com/science/article/pii/S0306919217310552
Bohnes, FA; Hauschild, MZ; Schlundt, J; Laurent, A. (2019). Life cycle assessments of aquaculture systems: A
critical review of reported findings with recommendations for policy and system development. Reviews in
Aquaculture 11: 1061 -1079. https://doi.org/10.1111/rag. 12280
Bozeman, JF, III; Ashton, WS; Theis, TL. (2019). Distinguishing environmental impacts of household food-
spending patterns among U.S. demographic groups. Environmental Engineering Science 36: 763-777.
https://doi.Org/10.1089/ees.2018.0433
Buzby, JC; Farah-Wells, H; Hyman, J. (2014). The estimated amount, value, and calories of postharvest food
losses at the retail and consumer levels in the United States. USDA-ERS Economic Information Bulletin.
Canning, P; Rehkamp, S; Hitaj, C; Peters, C. (2020). Resource requirements of food demand in the United
States. (ERR 273). U.S. Department of Agriculture, Economic Research Service.
https://www.ers.usda.gov/webdocs/publications/98401 Zerr-273.pdf? v=9205.2
Capel, PD; McCarthy, KA; Coupe, RH; Grey, KM; Amenumey, SE; Baker, NT; Johnson, RL. (2018). Agriculture —
A river runs through it — The connections between agriculture and water quality [Report], In Circular (pp.
1-201). (1433). Reston, VA. http://pubs.er.usgs.gov/publication/cir1433
Cattaneo, A; Federighi, G; Vaz, S. (2020). The environmental impact of reducing food loss and waste: A critical
assessment. Food Policy101890. http://www.sciencedirect.com/science/article/pii/S030691922030092Q
CEC (Commission for Environmental Cooperation). (2017). Characterization and management of food loss and
waste in North America. Montreal, Canada. http://www3.cec.Org/islandora/en/item/11772-
characterization-and-management-food-loss-and-waste-in-north-america-en.pdf
Champions 12.3. (2017a). The business case for reducing food loss and waste.
https://champions123.org/sites/default/files/2020-08/business-case-for-reducing-food-loss-and-waste.pdf
Champions 12.3. (2017b). Road map to achieving SDG target 12.3.
https://champions123.org/publication/roadmap-achieving-sdg-target-123
Chaudhary, A; Brooks, TM. (2019). National Consumption and Global Trade Impacts on Biodiversity. World
Development 121: 178-187. https://www.sciencedirect.com/science/article/pii/S03Q5750X173Q3261
References
84
-------
Chaudhary, A; Kastner, T. (2016). Land use biodiversity impacts embodied in international food trade. Global
Environmental Change 38: 195-204.
https://www.sciencedirect.com/science/article/pii/S0959378Q1630Q346
Chen, C; Chaudhary, A; Mathys, A. (2020). Nutritional and environmental losses embedded in global food waste.
Resources, Conservation and Recycling 160: 104912. https://doi.orq/10.1016/i.resconrec.2020.104912
Chen, HS. (2019). Environmental Concerns and Food Consumption: What Drives Consumers' Actions to Reduce
Food Waste? Journal of International Food & Agribusiness Marketing 31: 273-292.
httPs://doi.ora/10.1080/08974438.2018.1520179
Clark, M; Tilman, D. (2017). Comparative analysis of environmental impacts of agricultural production systems,
agricultural input efficiency, and food choice. Environmental Research Letters 12: 064016.
https://doi.ora/10.1088/1748-9326/aa6cd5
Clark, MA; Domingo, NGG; Colgan, K; Thakrar, SK; Tilman, D; Lynch, J; Azevedo, IL; Hill, JD. (2020). Global
food system emissions could preclude achieving the 1.5° and 2°C climate change targets. Science 370:
705. https://doi.orq/10.1126/science.aba7357
Compton, J. (2021). Cost of N20 release from human activities to the environment. International Nitrogen
Initiative, Social Cost of Nitrous Oxide (N20).
https://cfpub.epa.gov/si/si public record Report.cfm?dirEntrvld=350908&Lab=CPHEA
Conijn, JG; Bindraban, PS; Schroder, JJ; Jongschaap, REE. (2018). Can our global food system meet food
demand within planetary boundaries? Agriculture, Ecosystems & Environment 251: 244-256.
https://doi.org/10.1016/i.agee.2017.06.001
Conrad, Z; Niles, MT; Neher, DA; Roy, ED; Tichenor, NE; Jahns, L. (2018). Relationship between food waste, diet
quality, and environmental sustainability. PloS One 13: e0195405-e0195405.
https://pubmed.ncbi.nlm.nih.gov/29668732
Coudard, A; Corbin, E; de Koning, J; Tukker, A; Mogollon, JM. (2021). Global water and energy losses from
consumer avoidable food waste. Journal of Cleaner Production 326: 129342.
https://www.sciencedirect.com/science/article/pii/S0959652621035265
Crippa, M; Solazzo, E; Guizzardi, D; Monforti-Ferrario, F; Tubiello, FN; Leip, A. (2021). Food systems are
responsible for a third of global anthropogenic GHG emissions. Nature Food 2: 198-209.
https://doi.ora/10.1038/s43016-021 -00225-9
CRS (Congressional Research Service). (2020). U.S. Food and Agricultural Imports: Safeguards and selected
issues.
https://crs re ports, congress. gov/product/pdf/R/R46440/2#:~:text=The%20U.S.%20Department%20of%20
Agriculture%20(USDA')%20reports%20that%20food%20imports.or%20product%20value%20of%20trade
Cuellar, AD; Webber, ME. (2010). Wasted food, wasted energy: The embedded energy in food waste in the
United States. Environmental Science & Technology 44: 6464-6469. https://doi.org/10.1021/es10031 Od
D'Odorico, P; Davis, KF; Rosa, L; Carr, JA; Chiarelli, D; Dell'Angelo, J; Gephart, J; MacDonald, GK; Seekell, DA;
Suweis, S; Rulli, MC. (2018). The global food-energy-water nexus. Reviews of Geophysics 56: 456-531.
https://doi.org/10.1029/2017RG000591
Danish EPA (Danish Environmental Protection Agency). (2018). Kortlasgning af sammenstaetningen af
dagrenovation og kildesorteret oranisk affald fra husholdninger.
https://www2.mst.dk/Udgiv/publikationer/2018/03/978-87-93614-78-9.pdf
Davidson, E. (2009). The contribution of manure and fertilizer nitrogen to atmospheric nitrous oxide since 1860.
Nature Geoscience 2: 659-662. https://doi.org/10.1038/ngeo608
Diaz, RJ; Rosenberg, R. (2008). Spreading Dead Zones and Consequences for Marine Ecosystems. Science
321: 926-929. https://doi.oro/10.1126/science.1156401
Dieter, CA; Maupin, MA; Caldwell, RR; Harris, MA; Ivahnenko, Tl; Lovelace, JK; Barber, NL; Linsey, KS. (2018).
Estimated use of water in the United States in 2015 [Report], In Circular (pp. 76). (1441). Reston, VA:
USGS. http://pubs.er.usgs.gov/publication/cir1441
Distefano, T; Kelly, S. (2017). Are we in deep water? Water scarcity and its limits to economic growth. Ecological
Economics 142: 130-147. https://www.sciencedirect.com/science/article/pii/S0921800916310795
Dupigny-Giroux, L; E.L. Mecray, E; Lemcke-Stampone, M; Hodgkins, G; Lentz, E; Mills, K; Lane, E; Miller, R;
Hollinger, D; Solecki, W; Wellenius, G; Sheffield, P; MacDonald, A; Caldwell, C. (2018). Impacts, Risks,
and Adaptation in the United States. In Fourth National Climate Assessment. Washington, DC: U.S.
Global Change Research Program. https://nca2018.globalchange.gov/chapter/1/
EIA (U.S. Energy Information Administration). (2020). 2019 Annual Energy Outlook, Table A4: Residential Sector
Key Indicators and Consumption.
Eshel, G; Shepon, A; Makov, T; Milo, R. (2014). Land, irrigation water, greenhouse gas, and reactive nitrogen
burdens of meat, eggs, and dairy production in the United States. Proceedings of the National Academy
of Sciences of the United States of America 111: 11996-12001.
References
85
-------
FAO (Food and Agriculture Organization of the United Nations). (2011). Global food losses and waste. Food and
Agriculture Organization of the United Nations, http://www.fao.orq/docrep/014/mb060e/mb060e00.pdf
FAO (Food and Agriculture Organization of the United Nations). (2013a). Food wastage footprint: Impacts on
natural resources: Summary report. Food and Agriculture Organization of the United Nations.
http://www.fao.Org/3/i3347e/i3347e.pdf
FAO. (2013b). Food wastage footprint: Impacts on natural resources: Technical report. Food and Agriculture
Organization of the United Nations. https://www.fao.Org/3/ar429e/ar429e.pdf
FAO (Food and Agriculture Organization of the United Nations). (2015a). AQUASTAT methodology - Water use.
Food and Agriculture Organization of the United Nations.
http://www.fao.org/aguastat/en/overview/methodology/water-use
FAO (Food and Agriculture Organization of the United Nations). (2015b). Food wastage footprint & Climate
Change. Food and Agriculture Organization of the United Nations.
http://www.fao.Org/3/bb144e/bb144e.pdf
FAO (Food and Agriculture Organization of the United Nations). (2017). FAOSTAT: New food balances. Food and
Agriculture Organization of the United Nations. http://www.fao.Org/faostat/en/#data/FBS/metadata
FAO (Food and Agriculture Organization of the United Nations). (2018). FAOSTAT. Food and Agriculture
Organization of the United Nations. http://www.fao.Org/faostat/en/#countrv/231
FAO (Food and Agriculture Organization of the United Nations). (2019a). Food loss index. Online statistical
working system for loss calculations. Food and Agriculture Organization of the United Nations.
http://www.fao.org/platform-food-loss-waste/flw-data/en/
FAO (Food and Agriculture Organization of the United Nations). (2019b). The state of food and agriculture:
Moving forward on food loss and waste reduction. Rome, Italy: Food and Agriculture Organization of the
United Nations, http://www.fao.org/3/ca6030en/ca6030en.pdf
FAO (Food and Agriculture Organization of the United Nations). (2020). Sustainable Development Goals:
Indicator 12.3.1 - Global food loss and waste. Food and Agriculture Organization of the United Nations.
http://www.fao.org/sustainable-development-goals/indicators/1231/en/
FAO. (2021). Food Balances. Food and Agriculture Organization of the United Nations
https://www.fao.0rg/faostat/en/#data/FBS
Fernandez-Cornejo, J, Richard Nehring, Craig Osteen, Seth Wechsler, Andrew Martin, and Alex Vialou. (2014).
Pesticide Use in U.S. Agriculture: 21 Selected Crops, 1960-2008. U.S. Department of Agriculture,
Economic Research Service, https://papers.ssrn.com/sol3/papers.cfm7abstract id=2502986
Flanagan, K; Lipinski, B; Goodwin, L. (2019). SDG target 12.3 on food loss and waste: 2019 Progress report.
Champions 12.3. https://champions123.org/wp-content/uploads/2019/09/champions-12-3-2019-prooress-
report.pdf
Foley, JA; DeFries, R; Asner Gregory, P; Barford, C; Bonan, G; Carpenter Stephen, R; Chapin, FS; Coe Michael,
T; Daily Gretchen, C; Gibbs Holly, K; Helkowski Joseph, H; Holloway, T; Howard Erica, A; Kucharik
Christopher, J; Monfreda, C; Patz Jonathan, A; Prentice, IC; Ramankutty, N; Snyder Peter, K. (2005).
Global Consequences of Land Use. Science 309: 570-574. https://doi.Org/10.1126/science.1111772
Foley, JA; Ramankutty, N; Brauman, KA; Cassidy, ES; Gerber, JS; Johnston, M; Mueller, ND; O'Connell, C; Ray,
DK; West, PC; Balzer, C; Bennett, EM; Carpenter, SR; Hill, J; Monfreda, C; Polasky, S; Rockstrom, J;
Sheehan, J; Siebert, S; Tilman, D; Zaks, DPM. (2011). Solutions for a cultivated planet. Nature 478: 337-
342. https://d0i.0rg/l0.1038/nature10452
Fry, JP; Ceryes, CA; Voorhees, JM; Barnes, NA; Love, DC; Barnes, ME. (2019). Occupational safety and health
in U.S. aquaculture: A review. Journal of Agromedicine 24: 405-423.
httPs://doi.oro/10.1080/1059924X.2019.1639574
Garnett, T. (2011). Where are the best opportunities for reducing greenhouse gas emissions in the food system
(including the food chain)? Food Policy 36: S23-S32.
https://www.sciencedirect.com/science/article/pii/S0306919210001132
Gerten, D; Heck, V; Jagermeyr, J; Bodirsky, BL; Fetzer, I; Jalava, M; Kummu, M; Lucht, W; Rockstrom, J;
Schaphoff, S; Schellnhuber, HJ. (2020). Feeding ten billion people is possible within four terrestrial
planetary boundaries. Nature Sustainability 3: 200-208. https://doi.Org/10.1038/s41893-019-0465-1
Godfray, HCJ; Crute, IR; Haddad, L; Lawrence, D; Muir, JF; Nisbett, N; Pretty, J; Robinson, S; Toulmin, C;
Whiteley, R. (2010). The future of the global food system. Philosophical Transactions of the Royal Society
B: Biological Sciences 365: 2769-2777. https://doi.Org/10.1098/rstb.2010.0180
Goossens, Y; Berrens, P; Custers, K; Van Hemelryck, S; Kellens, K; Geeraerd, A. (2019). How origin, packaging
and seasonality determine the environmental impact of apples, magnified by food waste and losses. The
International Journal of Life Cycle Assessment 24: 667-687. https://doi.Org/10.1007/s11367-018-1522-0
Guo, X; Broeze, J; Groot, JJ; Axmann, H; Vollebregt, M. (2020). A worldwide hotspot analysis on food loss and
waste, associated greenhouse gas emissions, and protein losses. Sustainability 12: 7488.
httPs://doi.oro/10.3390/su12187488
References
86
-------
Gustavsson, J; Cederberg, C; Sonesson, U. (2011). Global food waste and food losses: extent, causes and
prevention. United Nations, Food and Agriculture Organization, https://www.diva-
portal.orq/smash/qet/diva2:944159/FULLTEXT01 .pdf
Gustavsson, J; Cederberg, C; Sonesson, U; Emanuelsson, A. (2013). The methodology of the FAO study: "Global
food waste and food losses: extent, causes and prevention". (SIK report No. 857). Swedish Institute for
Food and Biotechnology.
Hall, KD; Guo, J; Dore, M; Chow, CC. (2009). The progressive increase of food waste in America and its
environmental impact. PLoS One 4: e7940-e7940. https://pubmed.ncbi.nlm.nih.gov/19946359
Hayek, MN; Harwatt, H; Ripple, WJ; Mueller, ND. (2021). The carbon opportunity cost of animal-sourced food
production on land. Nature Sustainability 4: 21-24. https://doi.org/10.1038/s41893-020-00603-4
Heard, BR; Bandekar, M; Vassa, B; Miller, SA. (2019). Comparison of life cycle environmental impacts from meal
kits and grocery store meals. Resources Conservation and Recycling 147: 189-200.
https://doi.Org/10.1016/i.resconrec.2019.04.008
Heller, MC. (2019). Waste not, want not: Reducing food loss and waste in North America through life cycle-based
approaches. Washington, DC: United Nations Environment Programme, North America Office.
Heller, MC; Keoleian, GA. (2015). Greenhouse gas emission estimates of U.S. dietary choices and food loss.
Journal of Industrial Ecology 19: 391-401. https://doi.Org/10.1111 /iiec. 12174
Heller, MC; Wllits-Smith, A; Meyer, R; Keoleian, GA; Rose, D. (2018). Greenhouse gas emissions and energy
use associated with production of individual self-selected US diets. Environmental Research Letters 13:
044004. http://dx.doi.oro/10.1088/1748-9326/aabOac
Hellerstein, D; Vilorio, D; Ribaudo, M (U.S. Department of Agriculture, Economic Research Service). (2019).
Agricultural Resources and Environmental Indicators, 2019.
https://www.ers.usda.gov/webdocs/publications/93026/eib-208.pdf
Hi
-------
Lopez Barrera, E; Hertel, T. (2021). Global food waste across the income spectrum: Implications for food prices,
production and resource use. Food Policy 98: 101874.
https://www.sciencedirect.com/science/article/pii/S030691922030Q762
Love, DC; Fry, JP; Milli, MC; Neff, RA. (2015). Wasted seafood in the United States: Quantifying loss from
production to consumption and moving toward solutions. Global Environmental Change 35: 116-124.
http://www.sciencedirect.com/science/article/pii/S095937801530034Q
Lundqvist, J; de Fraiture, C; Molden, D. (2008). Saving Water: From Field to Fork - Curbing Losses and Wastage
in the Food Chain. In SIWI Policy Brief. SIWI.
Marston, L; Ao, Y; Konar, M; Mekonnen, M; Hoekstra, A. (2018). High-resolution water footprints of production of
the United States. Water Resources Research 54: 2288-2316.
Marston, LT; Read, QD; Brown, SP; Muth, MK. (2021). Reducing Water Scarcity by Reducing Food Loss and
Waste [Policy and Practice Reviews], Frontiers in Sustainable Food Systems 5.
https://www.frontiersin.org/article/10.3389/fsufs.2021.651476
Maryland DOE. (2021). Food Scraps Management. Available online at
https://mde.marvland.qov/proqrams/land/recvclinqandoperationsproqram/paqes/foodscraps.aspx
(accessed 10-06).
Mbow, C; Rosenzweig, C; Barioni, L; Benton, T; Herrero, M; Krishnapillai, M; Liwenga, E; Pradhan, P; Rivera-
Ferre, M; Sapkota, T; Tubiello, F; Xu, Y. (2019). Food Security. In Climate Change and Land: An IPCC
special report on climate change, desertification, land degradation, sustainable land management, food
security, and greenhouse gas fluxes in terrestrial ecosystems: Intergovernmental Panel on Climate
Change. https://www.ipcc.ch/site/assets/uploads/sites/4/2020/02/SRCCL-Chapter-5.pdf
McDermott, C; Elliott, D; Moreno, L; Brodersen, R; Mulder, C. (2019). Oregon wasted food study: Summary of
findings. Oregon Department of Environmental Quality.
https://www.oreqon.qov/deq/mm/Documents/ORWastedFoodMeasStudvSummarv.pdf
Mekonnen, MM; Fulton, J. (2018). The effect of diet changes and food loss reduction in reducing the water
footprint of an average American. Water International 43: 860-870.
https://doi.ora/10.1080/02508060.2018.1515571
Mekonnen, MM; Hoekstra, AY. (2011). The green, blue and grey water footprint of crops and derived crop
products. Hydrol Earth Syst Sci 15: 1577-1600. https://hess.copernicus.Org/articles/15/1577/2011 /
Mekonnen, MM; Hoekstra, AY. (2012). A Global Assessment of the Water Footprint of Farm Animal Products.
Ecosystems 15: 401-415. httPs://doi.oro/10.1007/s10021-011-9517-8
Merrill, D; Leatherby, L. (2018). Here's how America uses its land. Bloomberg.
https://www.bloomberq.com/oraphics/2018-us-land-use/ (accessed March 3, 2020).
Miller, R; Blair, P. (2009). Input-Output Analysis: Foundations and Extensions: Cambridge University Press.
Mohareb, EA; Heller, MC; Guthrie, PM. (2018). Cities' role in mitigating United States food system greenhouse
gas emissions. Environmental Science & Technology 52: 5545-5554.
https://doi.orq/10.1021/acs.est.7b02600
Molden, D. (2007). Water for food, water for life : a comprehensive assessment of water management in
agriculture. https://www.routledqe.com/Water-for-Food-Water-for-Life-A-Comprehensive-Assessment-of-
Water-Manaqement/Molden/p/book/9781844073962
Moreno, L. (2020). Is this edible? BioCycle. https://www.biocvcle.net/is-this-edible/
Muth, MK; Birney, C; Cuellar, A; Finn, SM; Freeman, M; Galloway, JN; Gee, I; Gephart, J; Jones, K; Low, L;
Meyer, E; Read, Q; Smith, T; Weitz, K; Zoubek, S. (2019). A systems approach to assessing
environmental and economic effects of food loss and waste interventions in the United States. Science of
the Total Environment 685: 1240-1254. https://doi.oro/10.1016/i.scitotenv.2019.06.230
NASEM (National Academies of Sciences, Engineering, Medicine). (2020). A National Strategy to Reduce Food
Waste at the Consumer Level. Washington, DC: The National Academies Press.
https://www.nap.edu/cataloo/25876/a-national-strateov-to-reduce-food-waste-at-the-consumer-level
NCSL (National Conference of State Legislatures). (2020). Fighting food waste.
https://www.ncsl.orq/research/environment-and-natural-resources/fiqhtinq-food-waste.aspx
Neff, RA; Spiker, ML; Truant, PL. (2015). Wasted Food: U.S. Consumers' Reported Awareness, Attitudes, and
Behaviors. PLoS One 10: e0127881. https://doi.org/10.1371/iournal.pone.0127881
Niles, MT; Ahuja, R; Barker, T; Esquivel, J; Gutterman, S; Heller, MC; Mango, N; Portner, D; Raimond, R; Tirado,
C; Vermeulen, S. (2018). Climate change mitigation beyond agriculture: A review of food system
opportunities and implications. Renewable Agriculture and Food Systems 33: 297-308.
https://www.cambridge.org/core/article/climate-change-mitigation-bevond-agriculture-a-review-of-food-
svstem-opportunities-and-implications/1A441CC574E74E8F29FEBD284CE92945
NOAA (National Oceanic and Atmospheric Administration). (2021). Global Aquaculture: Farmed Seafood Imports
to United States. Available online at https://www.fishwatch.gov/sustainable-seafood/the-global-
References
88
-------
picture#:~:text=Farmed%20Seafood%20lmports%20to%20United.of%20more%20than%20%2410.4%20
billion.
Our World in Data. (2021). Fertilizer use per hectare of cropland, 2002 to 2015.
https://ourworldindata.orq/qrapher/fertilizer-use-in-kq-per-hectare-of-arable-land?tab=chart
Pagani, M; De Menna, F; Johnson, TG; Vittuari, M. (2020). Impacts and costs of embodied and nutritional energy
of food losses in the US food system: farming and processing (Part A). Journal of Cleaner Production
244: 118730. http://www.sciencedirect.com/science/article/pii/S09596526193360Q5
Parker, RWR; Blanchard, JL; Gardner, C; Green, BS; Hartmann, K; Tyedmers, PH; Watson, RA. (2018). Fuel use
and greenhouse gas emissions of world fisheries. Nature Climate Change 8: 333-337.
httPs://doi.ora/10.1038/s41558-018-0117-x
Parry, A; Bleazard, P; Okawa, K. (2015). Preventing Food Waste: Case Studies of Japan and the United
Kingdom. Paris, France, https://www.oecd-ilibrarv.org/content/paper/5is4w29cf0f7-en
Peters, CJ; Picardy, J; Darrouzet-Nardi, AF; Wlkins, JL; Griffin, TS; Fick, GW. (2016). Carrying capacity of U.S.
agricultural land: Ten diet scenarios. Elementa: Science of the Anthropocene 4.
https://doi.orq/10.12952/iournal.elementa.000116
Pfister, S; Bayer, P; Koehler, A; Hellweg, S. (2011). Environmental Impacts of Water Use in Global Crop
Production: Hotspots and Trade-Offs with Land Use. Environmental Science & Technology 45: 5761-
5768. https://doi.ora/10.1021/es1041755
Poore, J; Nemecek, T. (2018). Reducing food's environmental impacts through producers and consumers.
Science 360: 987. http://science.sciencemag.org/content/360/6392/987.abstract
Porter, SD; Reay, DS; Higgins, P; Bomberg, E. (2016). A half-century of production-phase greenhouse gas
emissions from food loss & waste in the global food supply chain. Science of the Total Environment 571:
721-729. http://www.sciencedirect.com/science/article/pii/S0048969716314863
Power, AG. (2010). Ecosystem services and agriculture: tradeoffs and synergies. Philosophical Transactions of
the Royal Society B: Biological Sciences 365: 2959-2971. https://doi.Org/10.1098/rstb.2010.0143
Ramankutty, N; Evan, AT; Monfreda, C; Foley, JA. (2008). Farming the planet: 1. Geographic distribution of
global agricultural lands in the year 2000. Global Biogeochemical Cycles 22.
https://doi.Org/10.1029/2007GB002952
Ray, DK; Mueller, ND; West, PC; Foley, JA. (2013). Yield Trends Are Insufficient to Double Global Crop
Production by 2050. PLOS ONE 8: e66428. https://doi.org/10.1371/iournal.pone.0066428
Read, QD; Brown, S; Cuellar, AD; Finn, SM; Gephart, JA; Marston, LT; Meyer, E; Weitz, KA; Muth, MK. (2020).
Assessing the environmental impacts of halving food loss and waste along the food supply chain. Science
of the Total Environment 712: 136255. https://doi.Org/10.1016/i.scitotenv.2019.136255
ReFED. (2021a). Insights Engine Food Waste Monitor, https://insiohts-engine.refed.com/food-waste-
monitor?view=overview&vear=2019
ReFED. (2021b). U.S. Food Waste Policy Finder. Available online at https://policvfinder.refed.org/
Rehkamp, S; Canning, P. (2018). Measuring Embodied Blue Water in American Diets: An EIO Supply Chain
Approach. Ecological Economics 147: 179-188.
https://www.sciencedirect.com/science/article/pii/S09218009173Q455X
Ritchie, H; Roser, M. (2020). Environmental impact of food production. Our World in Data.
https://ourworldindata.org/environmental-impacts-of-food
Roos, E; Bajzelj, B; Smith, P; Patel, M; Little, D; Garnett, T. (2017). Greedy or needy? Land use and climate
impacts of food in 2050 under different livestock futures. Global Environmental Change 47: 1-12.
https://www.sciencedirect.com/science/article/pii/S0959378016306872
Rosenzweig, C; N Tubiello, F; Sandalow, D; Benoit, P; N Hayek, M. (2021). Finding and fixing food system
emissions: the double helix of science and policy. Environmental Research Letters 16: 061002.
http://dx.doi.oro/10.1088/1748-9326/ac0134
Roser, M; Ritchie, H. (2013). Food Supply. Our World in Data, https://ourworldindata.org/food-supplv
Rost, S; Gerten, D; Bondeau, A; Lucht, W; Rohwer, J; Schaphoff, S. (2008). Agricultural green and blue water
consumption and its influence on the global water system. Water Resources Research 44.
https://doi.Org/10.1029/2007WR006331
Rotz, CA; Asem-Hiablie, S; Place, S; Thoma, G. (2019). Environmental footprints of beef cattle production in the
United States. Agricultural Systems 169: 1-13.
http://www.sciencedirect.com/science/article/pii/S0308521X18305675
Searchinger, T; Waite, R; Hanson, C; Ranganathan, J; Dumas, P; Matthews, E; Klirs, C. (2019). Creating a
sustainable food future: A menu of solutions to feed nearly 10 billion people by 2050. World Resources
Institute, https://files.wri.org/s3fs-public/creating-sustainable-food-future 2.pdf
Skaf, L; Franzese, PP; Capone, R; Buonocore, E. (2021). Unfolding hidden environmental impacts of food waste:
An assessment for fifteen countries of the world. Journal of Cleaner Production 310: 127523.
https://www.sciencedirect.com/science/article/pii/S095965262101742X
References
89
-------
Smith, LP; Ng, SW; Popkin, BM. (2013). Trends in US home food preparation and consumption: analysis of
national nutrition surveys and time use studies from 1965-1966 to 2007-2008. Nutrition Journal 12: 45.
https://doi.ora/10.1186/1475-2891 -12-45
Smith, MR; Micha, R; Golden, CD; Mozaffarian, D; Myers, SS. (2016). Global Expanded Nutrient Supply (GENuS)
Model: A New Method for Estimating the Global Dietary Supply of Nutrients. PLOS ONE 11: e0146976.
https://doi.org/10.1371/iournal.pone.0146976
Spang, ES; Moreno, LC; Pace, SA; Achmon, Y; Donis-Gonzalez, I; Gosliner, WA; Jablonski-Sheffield, MP;
Momin, MA; Quested, TE; Winans, KS; Tomich, TP. (2019). Food loss and waste: Measurement, drivers,
and solutions. Annual Review of Environment and Resources 44: 117-156.
https://doi.ora/10.1146/annurev-environ-101718-033228
Spiker, ML; Hiza, HAB; Siddiqi, SM; Neff, RA. (2017). Wasted food, wasted nutrients: Nutrient loss from wasted
food in the United States and comparison to gaps in dietary intake. Journal of the Academy of Nutrition
and Dietetics 117: 1031-1040. https://doi.Org/10.1016/i.iand.2017.03.015
Springmann, M; Clark, M; Mason-D'Croz, D; Wiebe, K; Bodirsky, BL; Lassaletta, L; de Vries, W; Vermeulen, SJ;
Herrero, M; Carlson, KM; Jonell, M; Troell, M; DeClerck, F; Gordon, LJ; Zurayk, R; Scarborough, P;
Rayner, M; Loken, B; Fanzo, J; Godfray, HCJ; Tilman, D; Rockstrom, J; Willett, W. (2018). Options for
keeping the food system within environmental limits. Nature 562: 519-525.
https://doi.org/10.1038/s41586-018-0594-0
State of New Jersey. (2017). PL 136. Senate No. 3027.§§1,2 - C.13:1E-226 to 13:1E-227.
https://www.ni.gov/dep/dshw/food-waste/pl 2017 136.pdf
State of Oregon. (2020). Executive order No. 20-04 directing state agencies to take actions to reduce and
regulate greenhouse gas emissions. State of Oregon, Office of the Governor.
https://www.oregon.gov/gov/Documents/executive orders/eo 20-04.pdf
Stensgard, A; Hanssen, O. (2016). Food Waste in Norway 2010-2015: Final Report from the ForMat Project
https://ec.europa.eu/food/sites/food/files/safetv/docs/fw lib format-rapport-2016-eng.pdf
Tilman, D; Clark, M. (2014). Global diets link environmental sustainability and human health. Nature 515: 518-
522. https://doi.Org/10.1038/nature13959
Tilman, D; Clark, M; Williams, DR; Kimmel, K; Polasky, S; Packer, C. (2017). Future threats to biodiversity and
pathways to their prevention. Nature 546: 73-81. https://doi.Org/10.1038/nature22900
Tlusty, MF; Tyedmers, P; Bailey, M; Ziegler, F; Henriksson, PJG; Bene, C; Bush, S; Newton, R; Asche, F; Little,
DC; Troell, M; Jonell, M. (2019). Reframing the sustainable seafood narrative. Global Environmental
Change 59: 101991. http://www.sciencedirect.com/science/article/pii/S0959378018313530
Tom, MS; Fischbeck, PS; Hendrickson, CT. (2016). Energy use, blue water footprint, and greenhouse gas
emissions for current food consumption patterns and dietary recommendations in the US. Environment
Systems and Decisions 36: 92-103. https://doi.Org/10.1007/s10669-015-9577-v
Toth, JD; Dou, Z. (2016). Wasted food, wasted resources: Land, irrigation water, and nutrients associated with
food wastage in the U.S. In Food Waste Across the Supply Chain: A US Perspective on Global Problem.
Ames, Iowa: Council for Agricultural Science and Technology, https://www.cast-science.org/wp-
content/uploads/2016/03/CAST-Food-Waste-Across-the-Supplv-Chain-2016.pdf
U.S. DoE (U.S. Department of Energy). (2020). Frequently asked questions (FAQS).
https://www.eia. gov/tools/fags/fag.php?id=97&t=3
U.S. EPA (U.S. Environmental Protection Agency). (2005). Protecting Water Quality from Agricultural Runoff.
Washington, DC. https://www.epa.gov/sites/production/files/2015-09/documents/ag runoff fact sheet.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2015). Mississippi river/Gulf of Mexico watershed nutrient
task force: 2015 Report to congress, https://www.epa.gov/sites/production/files/2015-
10/documents/htf report to congress final - 10.1.15.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2018). WaterSense. https://www.epa.gov/watersense/how-
we-use-water
U.S. EPA (U.S. Environmental Protection Agency). (2019a). Fertilizer Applied for Agricultural Purposes.
https://cfpub.epa.Qov/roe/indicator.cfm?i=55#1
U.S. EPA (U.S. Environmental Protection Agency). (2019b). Inventory of U.S. greenhouse gas emissions and
sinks 1990 - 2017. (EPA 430-R-19-001). https://www.epa.gov/sites/production/files/2019-
04/documents/us-ghg-inventorv-2019-main-text.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2020a). 2018 Wasted Food Report.
https://www.epa.gov/sites/production/files/2020-11/documents/2018 wasted food report.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2020b). Food: Material-specific data.
https://www.epa.gov/facts-and-figures-about-materials-waste-and-recvcling/food-material-specific-data
U.S. EPA (U.S. Environmental Protection Agency). (2020c). Report on the Environment: Land.
https://www.epa.gov/report-environment/land
References
90
-------
U.S. EPA (U.S. Environmental Protection Agency). (2020d). United States food loss and waste 2030 champions.
https://www.epa.qov/sustainable-manaqement-food/united-states-food-loss-and-waste-2030-
champions#q7 (accessed August 14, 2020).
U.S. EPA (U.S. Environmental Protection Agency). (2020e). User Manual for the Sustainable Materials
Management Prioritization Tools, https://www.epa.gov/sites/production/files/2020-
02/documents/user manual for smm prioritization tools.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2020f). Wasted food measurement methodology scoping
memo, https://www.epa.oov/sites/production/files/2020-
06/documents/food measurement methodology scoping memo-6-18-20.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2021a). Energy and the Environment: Greenhouse Gas
Equivalencies Calculator, https://www.epa.gov/energy/greenhouse-gas-eguivalencies-calculator
U.S. EPA (U.S. Environmental Protection Agency). (2021b). Energy and the Environment: Greenhouse Gases
Equivalencies Calculator - Calculations and References, https://www.epa.gov/energy/greenhouse-gases-
eguivalencies-calculator-calculations-and-references
U.S. EPA (U.S. Environmental Protection Agency). (2021c). Inventory of U.S. Greenhouse Gas Emissions and
Sinks: 1990-2019. https://www.epa.gov/sites/production/files/2021-04/documents/us-ghg-inventorv-2Q21-
main-text, pdf
UN (United Nations). (2015). Transforming our world: The 2030 agenda for sustainable development.
(A/RES/70/1). United Nations General Assembly.
https://www.un.org/en/development/desa/population/migration/generalassemblv/docs/globalcompact/A R
ES 70 1 E.pdf
UN (United Nations). (2020a). 2019 Revision of World Population Prospects, https://population.un.org/wpp/
UN (United Nations). (2020b). Human Development Index (HDI). http://hdr.undp.org/en/content/human-
development-index-hdi
UN (United Nations ). (2021). SDG 12.3 Food waste index, https://www.unep.org/thinkeatsave/about/sdg-123-
food-waste-index
UNFCC (United Nations Framework Convention on Climate Change). (2015). The Paris Agreement.
https://unfccc.int/sites/default/files/english paris agreement.pdf
USCB. (2021). State Area Measurements and Internal Point Coordinates. Available online at
https://www.census.gov/geographies/reference-files/2010/geo/state-area.html
USDA (U.S. Department of Agriculture). (1996a). Soil Quality Indicators: Organic Matter.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcs142p2 053150.pdf
USDA (U.S. Department of Agriculture). (1996b). Soil Quality Resource Concerns: Compaction.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcs142p2 051594.pdf
USDA (U.S. Department of Agriculture). (1996c). Soil Quality Resource Concerns: Soil Erosion.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcs142p2 051278.pdf
USDA. (2012). What We Eat in America, NHANES 2009-2010 survey, https://www.ars.usda.gov/northeast-
area/beltsville-md-bhnrc/beltsville-human-nutrition-research-center/food-survevs-research-
group/docs/wweia-data-tables/
USDA (U.S. Department of Agriculture ). (2013). Farm and Ranch Irrigation Survey.
https://www.nass.usda.gov/Survevs/Guide to NASS Surveys/Farm and Ranch Irrigation/
USDA (U.S. Department of Agriculture). (2015a). Archived nutrient availability tables. U.S. Department of
Agriculture, Economic Research Service, https://www.ers.usda.gov/data-products/food-availabilitv-per-
capita-data-svstem/food-availabilitv-per-capita-data-
svstem/#Archived%20Nutrient%20Availabilitv%20Table
USDA (U.S. Department of Agriculture). (2015b). Census of Agriculture Survey Program - Environmental Sector.
https://www.nass.usda.gov/AgCensus/
USDA (U.S. Department of Agriculture). (2016). American diet includes many high-value imported products. U.S.
Department of Agriculture, Economic Research Service, https://www.ers.usda.gov/data-products/chart-
gallerv/gallerv/chart-detail/?chartld=58398
USDA (U.S. Department of Agriculture). (2017). Major uses of land in the United States, 1959-2012.
https://www.ers.usda.gov/webdocs/charts/55907/mlusummarv d.html?v=3909.7
USDA (U.S. Department of Agriculture). (2018). Nutrient Lists from Standard Reference Legacy (2018).
https://www.nal.usda.gov/fnic/nutrient-lists-standard-reference-legacv-2018
USDA (U.S. Department of Agriculture). (2019a). Data Products: Fertilizer Use and Price Workbook.
https://www.ers.usda.gov/webdocs/DataFiles/50341/fertilizeruse.xls?v=7031.1
USDA (U.S. Department of Agriculture). (2019b). Fertilizers & Pesticides, https://www.ers.usda.gov/topics/farm-
practices-management/fertilizers-pesticides/
References
91
-------
USDA (U.S. Department of Agriculture). (2019c). Food Availability and Consumption. U.S. Department of
Agriculture, Economic Research Service, https://www.ers.usda.gov/data-products/ag-and-food-statistics-
chartinq-the-essentials/food-availabilitv-and-consumption/
USDA (U.S. Department of Agriculture). (2019d). Food consumption and nutrient intakes. U.S. Department of
Agriculture, Economic Research Service, https://www.ers.usda.gov/data-products/food-consumption-and-
nutrient-intakes/ (accessed January 6, 2020).
USDA (U.S. Department of Agriculture). (2019e). Irrigation and water management. (ACH17-12). U.S.
Department of Agriculture, National Agricultural Statistics Service.
https://www.nass.usda.gov/Publications/Highlights/2019/2017Census Irrigation and WaterManaoement.
pdf
USDA (U.S. Department of Agriculture). (2019f). Nearly two-thirds of U.S. agricultural imports consist of
horticultural and tropical products. U.S. Department of Agriculture, Economic Research Service.
https://www.ers. usda.gov/data-products/chart-gallerv/gallerv/chart-detail/?chartld=58362
USDA. (2020). Summary data on annual food imports, values and volume by food category and source country,
1999-2017. U.S. Department of Agriculture, Economic Research Service, https://www.ers.usda.gov/data-
products/us-food-imports/
USDA (U.S. Department of Agriculture). (2021a). Food Security in the U.S.: Key Statistics & Graphics.
https://www.ers.usda.gov/topics/food-nutrition-assistance/food-securitv-in-the-us/kev-statistics-
graphics.aspx
USDA. (2021 b). Tracking the U.S. Domestic Food Supply Chain's Freshwater Use over Time.
USGS (U.S. Geological Survey). (2000). Pesticides in stream sediment and aquatic biota.
http://water.usgs.gov/nawga/pnsp/pubs/fs09200/
USGS (U.S. Geological Survey). (2019). Aquaculture Water Use. https://www.usos.gov/mission-areas/water-
resources/science/aguaculture-water-use?gt-science center obiects=Q#gt-science center objects
van den Bos Verma, M; de Vreede, L; Achterbosch, T; Rutten, MM. (2020). Consumers discard a lot more food
than widely believed: Estimates of global food waste using an energy gap approach and affluence
elasticity of food waste. PLoS One 15: e0228369. https://doi.org/10.1371/iournal.pone.0228369
Venkat, K. (2012). The climate change and economic impacts of food waste in the United States. International
Journal on Food System Dynamics 2. https://doi.org/10.18461/iifsd.v2i4.247
Vilarino, MV; Franco, C; Quarrington, C. (2017). Food loss and waste reduction as an integral part of a circular
economy [Mini Review], Frontiers in Environmental Science 5. https://doi.org/10.3389/fenvs.2017.00Q21
Vittuari, M; Pagani, M; Johnson, TG; De Menna, F. (2020). Impacts and costs of embodied and nutritional energy
of food waste in the US food system: Distribution and consumption (Part B). Journal of Cleaner
Production 252: 119857. http://www.sciencedirect.com/science/article/pii/S0959652619347274
Wakeland, W; S., C; K., V. (2012). Food transportation issues and reducing carbon footprint. In Green
Technologies in Food Production and Processing. Boston, MA: Springer. https://doi.Org/10.1007/978-1 -
4614-1587-9 9
Weber, CL; Matthews, HS. (2008). Food-miles and the relative climate impacts of food choices in the United
States. Environmental Science & Technology 42: 3508-3513. https://doi.org/10.1021/es702969f
Weil, R, and Nyle Brady. (2017). The Nature and Properties of Soil (15th ed.): Pearson Education.
https://www. pearson. com/store/p/elements-of-the-nature-and-properties-of-
soils/P100000860305/9780137504831 ?creative=545370970254&kevword=&matchtvpe=b&network=g&d
evice=c&gclid=CiOKCQiwlOmLBhCHARIsAGiJg7mDgKXQJDFoEvt97zYfvLV7JRI3Epb6VEJ FKoXFSQN
VbXILkARahcaAsxSEALw wcB&gclsrc=aw.ds
Wllett, W; Rockstrom, J; Loken, B; Springmann, M; Lang, T; Vermeulen, S; Garnett, T; Tilman, D; DeClerck, F;
Wood, A; Jonell, M; Clark, M; Gordon, LJ; Fanzo, J; Hawkes, C; Zurayk, R; Rivera, JA; De Vries, W;
Majele Sibanda, L; Afshin, A; Chaudhary, A; Herrero, M; Agustina, R; Branca, F; Lartey, A; Fan, S; Crona,
B; Fox, E; Bignet, V; Troell, M; Lindahl, T; Singh, S; Cornell, SE; Srinath Reddy, K; Narain, S; Nishtar, S;
Murray, CJL. (2019). Food in the Anthropocene: the EAT-Lancet Commission on healthy diets from
sustainable food systems. The Lancet 393: 447-492.
https://www.sciencedirect.com/science/article/pii/S0140673618317884
Wood, SLR; Alam, M; Dupras, J. (2019). Multiple pathways to more sustainable diets: Shifts in diet composition,
caloric intake and food waste. Frontiers in Sustainable Food Systems 3.
https://doi.oro/10.3389/fsufs.2019.00089
World Bank. (2021). United States Food Imports by country in US$ Thousand 2019.
https://wits.worldbank.org/CountrvProfile/en/Countrv/USA/Year/2019/TradeFlow/lmport/Partner/bv-
countrv/Product/Food
WRAP (Waste & Resources Action Programme). (2013). Household food and drink waste in the UK 2012.
https://wrap.org.uk/resources/report/household-food-and-drink-waste-united-kingdom-2012
References
92
-------
WRAP (Waste & Resources Action Programme). (2015). Strategies to achieve economic and environmental
gains by reducing food waste. https://newclimateeconomv.re port/workinqpapers/wp-
content/uploads/sites/5/2016/04/WRAP-NCE Economic-environmental-qains-food-waste.pdf
WRAP (Waste and Resources Action Programme). (2020). Food surplus and waste in the UK - key facts.
https://wrap.orq.Uk/resources/report/food-surplus-and-waste-uk-kev-facts#
Wuebbles, DJ; Fahey, DW; Hibbard, KA; DeAngelo, B; Doherty, S; Hayhoe, K; Horton, R; Kossin, JP; Taylor, PC;
Waple, AM; Weaver, CP. (2017). Executive summary. In Climate Science Special Report: Fourth National
Climate Assessment, Volume I. Washington, DC: U.S. Global Change Research Program.
https://science2017.olobalchanoe.gov/downloads/CSSR2017 PRINT Executive Summarv.pdf
WWF (World Wldlife Fund). (2018). No Food Left Behind: Part 1: Underutilized Produce Ripe for Alternative
Markets.
https://c402277.ssl.cf1 .rackcdn.com/publications/1170/files/oriqinal/WWF NoFoodLeftBehind820 2.pdf?1
564432069
Xue, L; Liu, G. (2019). Introduction to global food losses and food waste. In Saving Food: Production, Supply
Chain, Food Waste and Food Consumption: Academic Press. https://doi.orq/10.1016/B978-0-12-815357-
4.00001-8
Yang, Y; Ingwersen, WW; Hawkins, TR; Srocka, M; Meyer, DE. (2017). USEEIO: A new and transparent United
States environmentally-extended input-output model. Journal of Cleaner Production 158: 308-318.
https://www.sciencedirect.com/science/article/pii/S0959652617308806
Yu, Y; Jaenicke, EC. (2020). Estimating Food Waste as Household Production Inefficiency. American Journal of
Agricultural Economics 102.
References
93
-------
Appendix A: Inputs and Environmental Impacts
TABLE A-1. MAJOR INPUTS AND RESOURCES REQUIRED FOR U.S. CRADLE-TO-CONSUMER FOOD SYSTEM
Inputs
Food System
Stage
Category/
Type of Food
Water
Energy
(Electricity or
Fuels)
Land
Pesticides
Fertilizer
Sources
Primary
Production
Plants
(commodity
crops and
horticultural)
Irrigation
Farm equipment
fuel for planting,
fertilizing,
harvesting,
transportation
Planting
commodities
and
horticultural
and specialty
crops
Applied to many
commodities and
horticultural and
specialty crops
Applied to many
commodities and
horticultural and
specialty crops
D'Odorico et al.
(2018); Niles et
al. (2018); IOM
and NRC (2015)
Farm animals
Feed
production,
drinking
Feed production,
animal feeding,
animal housing,
manure handling,
transportation
Feed
production,
grazing, animal
housing
Applied to many
commodities used
for feed possibly
also to forage
plants
Applied to many
commodities
used for feed
possibly also to
forage plants
Asem-Hiablie et
al. (2019); Rotz et
al. (2019);
D'Odorico et al.
(2018); Niles et
al. (2018); IOM
and NRC (2015)
Seafood: wild
caught
Limited usage,
but requires
functional
aquatic
ecosystems
Boat fuel, cold
storage of caught
organisms
N/A
N/A
N/A
Niles et al.
(2018); Parker et
al. (2018); Avadi
Tapia et al.
(2016)
Seafood: farmed
Large water
storage
requirements
(both for
commercial and
feed organisms)
Boat fuel, and cold
storage of caught
organisms for feed,
electricity for
infrastructure and
maintenance (both
for commercial and
feed organisms)
Space for
infrastructure,
growth of
commercial and
feed organisms
Antibiotics and
anti-parasite
chemicals are
used in some
operations
N/A
Bohnes et al.
(2019); Fry et al.
(2019); Tlusty et
al. (2019); Niles
et al. (2018)
Appendix A: Inputs and Impacts of the U.S. Food System
94
-------
Inputs
Food System
Stage
Food
Category
Water
Energy
(Electricity or
Fuels)
Land
Pesticides
Fertilizer
Other
Sources
Handling and
Storage
Farm
animals
Drinking and
cleaning
Transportation
fuel, energy for
climatized
conditions for
animals, cold
storage for dairy
products
Infrastructure
for storage
and
transportation
N/A
N/A
N/A
Asem-Hiablie et
al. (2019); Niles
etal. (2018)
Plants
Washing some
products
Transportation
fuel, energy for
cold storage for
some horticultural
products
Infrastructure
for storage
and
transportation
N/A
N/A
N/A
Niles et al.
(2018)
Seafood:
wild caught
and farmed
N/A
Transportation
fuel, energy for
cold storage
Infrastructure
for storage
and for
transportation
N/A
N/A
N/A
Niles et al.
(2018)
Processing
and
Packaging
Farm
animals,
plants, and
seafood:
wild caught
and farmed
Food safety
and processing
needs
Energy for food
conversion from
raw materials to
final products
(e.g.,
slaughtering
animals for meat
and other
biproducts,
pasteurizing milk,
producing high-
fructose corn
syrup from corn,
filleting fish), cold
storage after
Infrastructure
for processing
and storage
N/A
N/A
Packaging
materials,
consumables,
chemicals
Asem-Hiablie et
al. (2019); Niles
etal. (2018)
Appendix A: Inputs and Impacts of the U.S. Food System
95
-------
Food System
Stage
Food
Category
Inputs
Water
Energy
(Electricity or
Fuels)
Land
Pesticides
Fertilizer
Other
Sources
processing prior
to distribution^)
Distribution
and Market
Farm
animals,
plants, and
seafood:
wild caught
and farmed
N/A
Transportation
fuel, energy for
cold storage for
some products
Infrastructure
for
transportation
N/A
N/A
N/A
Asem-Hiablie et
al. (2019); Niles
etal. (2018)
Consumption
Farm
animals,
plants, and
seafood:
wild caught
and farmed
Food
processing,
cooking and
food safety
needs
Energy for
refrigeration and
cooking, fuel for
transportation
N/A
N/A
N/A
N/A
Canning et al.
(2020); Pagani
et al. (2020);
Vittuari et al.
(2020); Asem-
Hiablie et al.
(2019); Niles et
al. (2018)
N/A = not applicable.
Appendix A: Inputs and Impacts of the U.S. Food System
96
-------
TABLE A-2. MAJOR ENVIRONMENTAL BURDENS ASSOCIATED WITH THE U.S. CRADLE-TO-CONSUMER FOOD SYSTEM
Food System
Stage
Category/
Impacts
Type of
Food
Biodiversity
Land
Water Use
Water Quality
Worker
Health
GHGs
Other Air
Sources
Primary
Plants
Conversion
Soil
Depletion of
Nitrogen and
Volatilization
N2O
Land
D'Odorico et
Production
(commodity
of land from
degradation
water
phosphorus
of pesticides
emissions
emissions of
al. (2018);
crops and
higher
, loss,
resources;
fertilizer runoff
and
from nitrogen
CO2,
Niles et al.
horticultural)
biodiversity to
compaction
reduced
of PO4-3, NO3-,
herbicides
fertilizer
methane,
(2018); IOM
plant
from tillage,
availability of
NH4+ leading to
during
application,
odors, fine
and NRC
production;
reduction of
ground and
water
application,
CO2
particulate
(2015)
can be
soil carbon
surface water
eutrophication
runoff of
emissions
matter
affected by
for other uses
pesticides
from farm
pesticide
into water,
equipment
application
nitrate
contaminatio
n of drinking
water
energy use,
loss of CO2
sequestration
Farm animals
Conversion
Soil
Depletion of
Manure
During
CH4 from
Manure and
Asem-Hiablie
of land from
degradation
water
management
application,
enteric
land
et al. (2019);
higher
, loss, and
resources;
runoff of PO4 3,
runoff of
fermentation
emissions of
Rotz et al.
biodiversity to
compaction
reduced
NO3-, NH4+
pesticides
and manure,
NOx, CO2,
(2019);
animal feed
from over
availability of
leading to
into water,
N2O from
ammonia,
D'Odorico et
and/or animal
grazing,
ground and
water
microbial
manure,
methane,
al. (2018);
production
leaching
surface water
eutrophication,
pathogens
pasture,
odors, fine
Niles et al.
from
for other uses
nitrate
from manure,
range and
particulate
(2018); IOM
manure
contamination
nitrate
cropland,
matter, ozone
and NRC
manageme
of drinking
contaminatio
CO2 from fuel
depletion
(2015)
nt
water
n of drinking
water
combustion
Seafood: wild
Overfishing
N/A
Minimal
Gear loss and
Minimal
GHG
Odors
Lebreton et
caught
can result in
biodiversity
collapse;
trawling can
damage sea
floor; gear
loss can
impacts
synthetic
fishing fibers
can result in
micro-plastic
accumulation
in marine
ecosystems
impacts
emissions
from energy
use for boat
fleets
al. (2018);
Niles et al.
(2018);
Parker et al.
(2018); Avadi
Tapia et al.
(2016)
Appendix A: Inputs and Impacts of the U.S. Food System
97
-------
Food System
Stage
Category/
Impacts
Type of
Food
Biodiversity
Land
Water Use
Water Quality
Worker
Health
GHGs
Other Air
Sources
result in
damage to
ecosystems
Seafood:
farmed
Ecotoxicity of
local
ecosystems
through
chemical use
and
introduction
of
nonindigenou
s species
Land
conversion
and
degradation
from
intensive
use
Intensive use
of water in
production
stage (i.e.,
growing
organisms)
Eutrophication
of water from
nitrogen and
phosphorus
from food
waste and
organism feces
Toxicity from
chlorine and
other
cleaning
products,
increased
potential for
disease
resistance
from
antibiotic use
GHG
emissions
from energy
use for
infrastructure,
equipment,
transportation
of feed and
materials
Odors
Bohnes et al.
(2019); Fry et
al. (2019);
Tlusty et al.
(2019); Niles
et al. (2018)
Appendix A: Inputs and Impacts of the U.S. Food System
98
-------
Food System
Stage
Food
Impacts
Category
Biodiversity
Land
Water Use
Water Quality
Worker
Health
GHGs
Other Air
Sources
Handling and
Storage
Farm animals
Minimal
impacts
Minimal
impacts
Depletion of
water
resources;
reduced
availability of
ground and
surface water
for other uses
Minimal
impacts
Minimal
impacts
GHG
emissions
from energy
use for
transportation
, climatized
conditions,
and cold
storage
Ozone
depletion,
refrigerant
leakage
Asem-Hiablie
etal. (2019);
Niles et al.
(2018)
Plants
Minimal
impacts
Minimal
impacts
Minimal
impacts
Minimal
impacts
Minimal
impacts
GHG
emissions
from energy
use for
transportation
, and cold
storage
Ozone
depletion,
refrigerant
leakage
Niles et al.
(2018)
Seafood: wild
caught and
farmed
Minimal
impacts
Minimal
impacts
Minimal
impacts
Minimal
impacts
Minimal
impacts
GHG
emissions
from energy
use for
transportation
, and cold
storage
Ozone
depletion,
refrigerant
leakage
Niles et al.
(2018)
Processing
and
Packaging
Farm
animals,
plants, and
seafood: wild
caught and
farmed
Minimal
impacts
Minimal
impacts
Depletion of
water
resources;
reduced
availability of
ground and
surface water
for other uses
Production of
wastewater
Minimal
impacts
GHG
emissions
from energy
for processing
Ozone
depletion,
volatile
organic
compounds,
refrigerant
leakage
Asem-Hiablie
etal. (2019);
Niles et al.
(2018)
Distribution
and market
Farm
animals,
plants, and
seafood: wild
Minimal
impacts
Minimal
impacts
Depletion of
water
resources;
reduced
availability of
Production of
wastewater
Minimal
impacts
GHG
emissions
from energy
use for
transportation
Ozone
depletion
Asem-Hiablie
etal. (2019);
Niles et al.
(2018)
Appendix A: Inputs and Impacts of the U.S. Food System
99
-------
Food System
Stage
Food
Category
Impacts
Biodiversity
Land
Water Use
Water Quality
Worker
Health
GHGs
Other Air
Sources
caught and
farmed
ground and
surface water
for other uses
, and cold
storage
Consumption
Farm
animals,
plants, and
seafood: wild
caught and
farmed
Minimal
impacts
Minimal
impacts
Depletion of
water
resources;
reduced
availability of
ground and
surface water
for other uses
Production of
wastewater
Minimal
impacts
GHG
emissions
from energy
use for
refrigeration
and cooking
and
transportation
Ozone
depletion
Canning et al.
(2020);
Pagani et al.
(2020);
Vittuari et al.
(2020); Asem-
Hiablie et al.
(2019); Niles
et al. (2018)
Appendix A: Inputs and Impacts of the U.S. Food System
100
-------
Appendix B: USDA and FAO FLW Data
Many of the FLW estimates presented in this report rely on data from USDA (Buzby et al., 2014's LAFA data) or
FAO (Gustavsson et al., 2011) for food availability data or food loss and waste rates, or both. For example,
Cuellar and Webber (2010), Venkat (2012), Heller and Keoleian (2015), Toth and Dou (2016), Spiker et al.
(2017), Birney et al. (2017), and Conrad et al. (2018) utilized USDA LAFA data. Whereas, CEC (2017), Chen et
al. (2020), Kummu et al. (2012) and Lipinski et al. (2013) relied on FAO/Gustavsson et al. (2011) food availability
and loss rates. This appendix provides more detail on the use of this data.
Several studies have relied on the USDA LAFA data and its loss rates. The loss rates within the LAFA data have
not been significantly updated since 2010 by Buzby et al. (2014). Consequently, researchers using differing years
of the LAFA data set have similar results. Cuellar and Webber (2010), Venkat (2012) and Heller and Keoleian
(2015) all used LAFA data without much additional manipulation and as expected, their per capita FLW estimates
are fairly aligned, 145, 180, 196 kg/capita/year, respectively. Mekonnen and Fulton (2018) used the 2015 LAFA
data resulting in an FLW estimate of 216 kg/capita/year and 1,237 daily calories per capita. Similarly, Spiker et al.
(2017) calculated the nutritional value of 2012 LAFA retail- and consumption-stage waste to be 1,217 daily
calories per capita which is close to the 2010 estimate by Buzby et al. (2014) of 1,249 daily calories per capita.
Toth and Dou (2016) and Birney et al. (2017) supplemented the LAFA data. Toth and Dou (2016) estimated FLW
from post-harvest, distribution and processing sector to be 36 million metric tons, based on their calculations
derived from the LAFA data, which combined with losses from the retail and consumption stages, equaled
approximately 100 million metric tons. Birney et al. (2017) didn't expand the stages included, but rather focused
on getting a more accurate estimate of the calories consumed, acknowledging that the Buzby et al. (2014) proxy
food consumption (which is food availability minus loss) estimate of 2,547 calories per capita per day overstates
what a typical American actually eats. Using the results of a study (Tom et al., 2016) which relied on USDA's
National Health and Nutrition Examination Survey (NHANES) data that calculates the average American's daily
calories, Birney et al. adjusted the LAFA food category calories to an estimated actual consumption of 2,390
calories per capita per day. The difference in food availability and actual consumption is considered FLW, hence
the higher overall FLW estimates.
NHANES, a nationwide survey conducted by the CDC, uses interviews and physical examination to collect
demographic and health information of the U.S. population. The dietary component of NHANES, known as What
We Eat in America (WWEIA), is a dietary recall survey of food eaten by about 5,000 individuals. Although other
sources of actual consumption are available from smaller studies of specific populations, NHANES is the largest
and most nationally representative source available. It is commonly cited for estimates of U.S. food consumption.
USDA's (2019c) analysis indicates mean consumption of 2,044 calories per capita per day from 2007 through
2010 of individuals 2 years old and older, with approximately 70 percent consumed at home and 30 percent away
from home.
Using the WWEIA data in NHANES, Conrad et al. (2018) quantified the amount of food consumed, identified its
comprised ingredients, and then back calculated how much of each ingredient needed to be produced given the
waste rates (using LAFA data for the portion of each ingredient that is lost or wasted at the consumer level). Since
Conrad et al. (2018) were focused on consumption stage waste, with a narrower scope, it follows that their FLW
estimates, by weight and calories, are lower than studies that included additional stages.
The FAO regional food loss and waste estimates developed by Gustavsson et al. (2011) and applied to the 2007
FAO Food Balance Sheet (FBS) data include edible FLW from production through the consumption stage for
eight categories of food at their primary commodity level. The methodology was described in greater detail by
Gustavsson et al. (2013). For the North America and Oceania region, Gustavsson et al. (2011) used a mix of U.S.
(predominantly from the USDA) and European data sources for the food loss and waste rates for each food
category within each stage of the food supply chain. Given the expansiveness of the scope and limited data,
some broad assumptions and extrapolations are used. For example, the loss rates for fruits and vegetables in
North America and Oceania are based on a study of carrots grown in Sweden and a statement about the
percentage of British-grown fruits and vegetables rejected by retailers (Gustavsson et al., 2013). Nonetheless, the
work by Gustavsson et al. (2011, 2013), captured the available information on FLW across all stages and enables
comparisons between global regions.
Appendix B: USDA and FAO FLW Data
101
-------
The FAO FLW data developed by Gustavsson et al. (2011, 2013) have been used by many studies, particularly
those making comparisons between countries and regions. The Commission for Environmental Cooperation
(CEC, 2017) used the 2007 FAO FBS data for the United States and Gustavsson et al. (2013) loss assumptions
for the NAO region, but included the inedible fractions of FLW (by excluding the conversion factors used by
Gustavsson et al.), which resulted in a higher per capita estimate. Also including inedible FLW but using more
recent food balance data and different loss rates, Guo et al. (2020) used the FAO FBS data for 2017 and applied
FLW rates from Porter et al. (2016). Read et al. (2020) relied on the Gustavsson et al. (2013) loss rates for North
America and Oceania to the USEEIO model. Since Gustavsson et al. (2011, 2013) doesn't include sweetners and
beverages, Read et al. (2020) supplemented the data set with Buzby et al. (2014) loss rate for sweetners, and
data from the United Kingdom WRAP program for loss estimates for beverages.
Lipinski et al (2013) applied the FLW estimated developed by Gustavsson et al. (2011, 2013) to the 2009 FAO
FBS data and converted the resulting food category weights into calories. Additionally, Kummu et al. (2012) also
converted the resulting FLW food category weights to calories. However, they excluded animal products, so their
per capita daily calories estimate is lower. Kummu et al. (2012) applied the Gustavsson et. al (2011, 2013) loss
rates to the FAO FBS data for vegetal products averaged over the years 2005-2007.
Taking a closer examination of the calories and nutrients in FLW at the consumption stage, Chen et al. (2020)
used the GENuS data set for the year 2011 for food availability (Smith et al., 2016) and applied the consumption
stage FLW rates from Gustavasson et al. (2011). In creating GENuS, Smith et al. (2016) disaggregated FAO's
FBS 94 food categories to 225, allowing for more detailed pairing with nutrition information. The resulting
estimates from Chen et al. (2020) are considerably lower than other studies using FAO data, which is expected
given that they were specifically examining the consumption stage waste. Even though the data sets differed, the
consumption stage food waste estimates from Chen et al. (2020) and Conrad et al. (2018) are within a similar
range, 184 kg and 154 kg per capita annually, respectively.
Appendix B: USDA and FAO FLW Data
102
-------
Appendix C: Literature Search Methodology
This appendix presents the literature search methodology used to identify, screen, and manage literature sources
for From Farm to Kitchen: The Environmental Impacts of U.S. Food Waste (Part 1) and associated issue papers.
The objective of this literature search was to identify the latest scientific information about food waste and food
waste reduction, including emerging technologies and approaches for prevention, reuse, and recycling. In
addition, analysis of the literature helped to identify knowledge gaps and the most important areas for future
scientific research.
Section B.1 describes the literature search methodology for peer-reviewed literature sources, and Section B.2
describes the identification of governmental and non-governmental reports that are not published in the peer-
reviewed scientific literature, referred to as "gray literature" in this methodology.
This literature search identified and prioritized 3,219 peer-reviewed sources, 1,723 of which were screened as
relevant to the scope of the From Farm to Kitchen report and issue papers. These source, as well as the key gray
literature (see Section B.2) and additional key sources identified in supplemental, targeted literature searches,
served as the primary corpus of literature from which literature synthesis and report development were performed.
The report and associated issue papers were developed by primarily using the literature identified through this
methodology, but were not limited to this set of literature as additional sources were identified subsequently (e.g.,
from peer-review recommendations).
C.1. Methodology for Peer-Reviewed Literature
Peer-reviewed literature was identified with a search of selected publication databases using keywords and
Boolean logic defined in this section. Titles and abstracts of the publications returned by the literature search were
processed to eliminate duplicates and then screened to identify a subset of "key" sources that meet criteria for
relevance and usefulness for the report or issue papers. Key sources were "tagged" to pre-defined topics to assist
authors in identifying the most relevant sources for particular topics covered in the report.
Peer-Reviewed Literature Search Strategy
The search of peer-reviewed literature focused on references relevant to the scope of the food waste report and
issue papers from 2010-present, with special priority given to more recent papers, which were considered to be
2017-present. A targeted search to identify review papers from 2014-present was performed. During
development of the report and issue papers, additional targeted searches were performed as needed within the
2010-present corpus of literature, and subject matter experts also identified key sources, some of which were
dated in 2020 or 2021.
The following databases were searched for relevant peer-reviewed literature:
• AGRICOLA (AGRICultural OnLine Access): AGRICOLA records describe publications and resources
encompassing all aspects of agriculture and allied disciplines, including animal and veterinary
sciences, entomology, plant sciences, forestry, aquaculture and fisheries, farming and farming
systems, agricultural economics, extension and education, food and human nutrition, and earth and
environmental sciences; Produced by the National Agricultural Library (NAL), U.S. Department of
Agriculture.
• AGRIS: AGRIS facilitates access to publications, journal articles, monographs, book chapters, and
grey literature - including unpublished science and technical reports, theses, dissertations and
conference papers in the area of agriculture and related sciences; Maintained by the Food and
Agriculture Organization of the United Nations (FAO).
• EBSCO: EBSCOhost Research Databases: Academic Search Complete; Energy & Power Source.
• PubMed: US National Library of Medicine National Institutes of Health.
• Web of Science: Web of Science Core Collection, refined by Research Area. Clarivate Analytics.
Appendix C: Literature Search Methodology
103
-------
Table A-1 outlines the searches performed and the combinations of keyword sets and Boolean operators used to
search each database. Four distinct sets of keywords were used to capture references with relevance to food
waste, pathways of food waste and food waste reduction, environmental impacts of food waste, and emerging
issues in the area of food waste. Sets were combined using Boolean logic to identify relevant references for
screening and evaluation. Search results were limited to publications written in English.
For each search, all references were downloaded into EndNote and then DeDuperwas used to remove duplicate
references (i.e., references that appeared in more than one of the databases searched). DeDuper is a tool that
uses a two-phase approach to identify and resolve duplicates: (1) it locates duplicates using automated logic, and
(2) it employs machine learning to predict likely duplicates which are then verified manually.
TABLE C-1. SEARCH STRATEGY KEYWORDS
Set
Search Keywords and Boolean Logic
Food Waste
Food AND (waste OR loss OR "FLW") AND (prevention OR system OR
consumed OR Surplus OR Excess OR Uneaten OR reduction OR supply OR
demand OR Per person OR Edible OR Inedible OR Safety OR recall OR
packaging OR Preventable OR Drivers OR Spoilage OR perishable OR
Freshness OR harvest OR transportation OR Processing OR manufacturing
OR supermarket OR grocer* OR reuse OR recycling OR seasonal OR
projection OR future OR economic)
Pathways
("Source reduction" OR Awareness OR education OR campaign OR
LeanPath OR Photodiary OR storage OR Labeling OR (Refrigerator AND
temperature) OR Cellar OR Frozen OR "Meal kits" OR packaging OR
Donation OR Upcycling OR "Animal feed" OR "Anaerobic digestion" OR Co-
digestion OR "Aerobic processes" OR Composting OR "Controlled
combustion" OR Incineration OR Landfill OR "Land application" OR de-
packaging OR "shelf life")
Environment
Environment* AND (use OR usage OR impacts) AND (climate OR "Air
emissions" OR "Water pollution" OR Pesticide OR Land OR Irrigation OR
Energy OR fertilizer OR water OR Herbicides))
Emerging
Issues
((Compost* or compostable) AND (packaging OR serviceware OR utensil OR
tableware OR plate OR bowl))
To efficiently screen results, references were prioritized using topic extraction, also referred to as clustering, with
ICF's Document Classification and Topic Extraction Resource (DoCTER) software. The titles and abstracts from
all search results (i.e., AGRICOLA, AGRIS, EBSCO, PubMed, and Web of Science) were run through DoCTER's
topic extraction function. Each study was assigned to a single cluster based on text similarities in titles and
abstracts. Clusters were prioritized or eliminated for screening based on the relevance of the keywords identified.
Only prioritized studies published from 2014-present were screened for relevance.
Peer-Reviewed Literature Screening and Tagging
The sources identified by the literature search were screened to identify those that are considered "key" sources
for the report and issue papers. To be considered a key source, a publication had to be relevant to the project
scope and exhibit at least most of the general attributes provided in EPA's Quality Assurance Instructions for
Contractors Citing Secondary Data, summarized below:
• Focus: the work not only addresses the area of inquiry under consideration but also contributes to its
understanding.
Appendix C: Literature Search Methodology
104
-------
• Verify: the work is consistent with accepted knowledge in the field or, if not, the new or varying
information is documented within the work; the work fits within the context of the literature and is
intellectually honest and authentic.
• Integrity: Is the work structurally sound? In a piece of research, is the design or research rationale
logical and appropriate?
• Rigor: the work is important, meaningful, and non-trivial relative to the field and exhibits enough depth
of intellect rather than superficial or simplistic reasoning.
• Utility: the work is useful and professionally relevant; it contributes to the field in terms of the
practitioners' understanding or decision-making on the topic.
• Clarity: Is it written clearly and appropriately for the nature of the study?
Relevance to the project scope was evaluated against the specific topics and criteria. In particular, relevant topics
included:
• Characterization of U.S. food waste, including but not limited to kinds of food, sources, amounts, and
reasons for loss or waste.
• Reduction strategies, including composting, anaerobic digestion, secondary industrial uses, animal
feed, donation, and source reduction.
• Lifecycle environmental costs and benefits of choices between and within levels of the EPA food
recovery hierarchy.
• Pre-processing technologies (e.g., grinding, heating, digestion) and their environmental implications
in use, including their potential to help reduce food waste.
• Food packaging and service ware and their relationships to food waste, including ways packaging
may impact prevention and recycling of food waste or use and value of products created by recycling.
• Chemical contaminants (e.g., PFOS, PFAS, persistent herbicides) and the risk and problems posed
in food waste streams.
• Food system trends to identify well-recognized trends in the U.S. food system that may impact food
waste and summarize what has been written about their potential impacts.
• Unharvested or unutilized crops that do not reach the consumer market.
• Waste or loss during transportation, food processing/manufacturing/packaging facilities, or wholesale
food distributors.
• Waste or loss at supermarkets (e.g., unsold or spoiled products), restaurants, and households.
• Existing economic, social, and cultural drivers of food waste or barriers to food waste prevention,
reuse, and recycling efforts.
The following topics were not considered relevant:
• Unutilized livestock (e.g., due to market forces, routine mortality) or unharvested or unutilized feed
crops.
• Regulatory drivers of food waste or barriers to food waste prevention, reuse, and recycling efforts.
• Broad economic impacts (e.g., on the agricultural sector) of food waste production, prevention, reuse,
and recycling efforts; economic costs and benefits for entities resulting from food waste production
and reduction strategies.
Appendix C: Literature Search Methodology
105
-------
The litstream™ tool was used to screen for key sources based on reference titles and abstracts, litstream™
facilitates screening by one or two independent reviewers, automatically compares categories, and identifies
discrepancies for resolution by another individual, litstream™ also allows users to design flexible data-extraction
forms, thus enabling the review team to perform the screening and tagging steps of the systematic review within
one software tool.
For publications identified as key sources, full text files were retrieved with EPA's Health & Environmental
Research Online (HERO) database as requested by authors. Then, authors used the full text of the key sources
to confirm topic area relevance and incorporate them into their literature synthesis.
A screening and tagging guidance document was developed to provide instructions and keywords associated with
the tags. To ensure internal consistency and accuracy of the litstream™ screening and tagging, a pilot screening
of 5-10 reference (per reviewer) was performed to provide feedback to the screening team. Additionally, 10% of
each reviewer's assigned citations were reviewed by a second reviewer. Discrepancies between the primary and
secondary reviews were resolved by lead authors.
C.2. Methodology for Grey Literature
Identifying key sources in the "grey literature" was essential to a comprehensive review and synthesis of the
report and issue papers. The review methodology for grey literature included a search strategy and approaches
for screening and tagging key sources.
Grey Literature Search Strategy
The peer-reviewed literature search was supplemented with relevant grey literature from the sources listed below:
• Grey literature publications cited by key sources identified by the EPA from prior related research.
These sources were screened as potential key sources.
• Grey literature publications identified by peer reviewers and subject matter experts who reviewed pre-
peer review drafts of the reports and issue papers (see the acknowledgments sections in the report
and each issue paper). These sources were considered key sources without screening.
• Targeted google and domain searches for selected governmental or non-governmental organizations.
The titles and URLs of potential sources identified by the searches were compiled in an Excel file used for
subsequent screening.
Grey Literature Screening and Tagging
Grey literature was screened in Excel using the key source criteria defined for peer-reviewed literature (see
Section A.1). Screeners applied the criteria to each of the potential sources in the database file described above
(i.e., titles and URLs identified from searches). For each URL, the screeners evaluated the sources by reviewing
abstracts, executive summaries, forewords, keyword lists, or tables of contents. When a screener identified a key
source, they recorded additional information including publishing organization, author names, and year for the
source to proceed to tagging.
Tagging was only performed for the grey literature identified as key sources, and the same tags as used for peer-
reviewed literature (see Section A.1) were used for grey literature, screeners applied the tags in columns within
Excel.
Appendix C: Literature Search Methodology
106
-------
v>EPA
U.S. Environmental Protection Agency
Office of Research and Development
------- |