Journal of Cleaner Production 279 (2021) 123722

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Review

Review of household food waste quantification methods: Focus on

composition analysis
Sohani Vihanga Withanage*, Goretty Maria Dias, Komal Habib
School of Environment, Enterprise and Development, University of Waterloo, ON, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Food loss and waste has become an increasingly discussed topic in recent years due to the associated
Received 11 March 2020 economic and environmental burden. Knowledge and methodological gaps in quantifying food waste
Received in revised form generation have been identified as the major obstacles in addressing the food waste generation issues by
29 July 2020
several researchers. Lack of standard methodology in quantifying food waste at households had led
Accepted 9 August 2020
Available online 22 August 2020
researchers to employ numerous methods that would generate incomparable results. Considering the
absence of a critical and comprehensive review of food waste quantification methods, the current study
Handling Editor: Prof. Jiri Jaromir Kleme
s aims at presenting a thorough literature review to compare household food waste quantification
methods with special focus on methods addressing composition analysis. In this review, a total of 45
studies considering four main food waste (FW) quantification methods, namely surveys, kitchen diaries,
waste audits and estimates based on secondary data are reviewed in detail to compare the strengths and
limitations of each method. The need for standardized methodologies for household food waste quan-
tification is further emphasized.
© 2020 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1. Aims of current study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1. Food loss and waste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.2. Impacts of food loss and waste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.3. Household FW and the gaps in existing knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1. Studies reviewed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.1. Analysis of bibliographic information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.1.2. Categorization of articles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.2. FW quantification and composition analysis methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.2.1. Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.2.2. Kitchen diaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2.3. Direct measurement through waste audits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.4. Estimates based on secondary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2.5. Mixed methods approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Declaration of competing interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Acknowledgement and funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

* Corresponding author.
E-mail addresses: [email protected], [email protected] (S.V. Withanage).

https://doi.org/10.1016/j.jclepro.2020.123722
0959-6526/© 2020 Elsevier Ltd. All rights reserved.
2 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

1. Introduction intended for human consumption frequently gets lost or wasted at

each stage of the food supply chain (FSC), from agricultural pro-
The current global population of 7 billion is estimated to reach 9.8 duction to final consumption, in other words, from farm to fork
billion people by the year 2050 (Godfray et al., 2010). With the (Xue et al., 2017). In general terms, this ‘food’ becomes ‘waste’ when
increasing population, the global demand for food will also escalate, it loses its quality or when it is not consumed by humans within the
imposing an inherent pressure on the global food supply system utility lifespan (FAO, 2011).
(Godfray et al., 2010). The Food and Agriculture Organization (FAO) There exist a number of definitions around food loss and food
has estimated that globally, 1.3 billion tonnes of food is being lost or waste. One of the widely accepted definitions are from FAO
wasted every year with an associated cost of 750 billion US dollars (2014a,b) where ‘food loss (FL)’ is defined as the “decrease in
(FAO, 2015). In Canada, it is estimated that 58% of total food pro- quality or quantity of food” that occurs at the initial production and
duction gets lost or wasted every year throughout the food supply distribution segments of the food supply chain (FSC) (FAO, 2014a).
chain. The monetary value of avoidable food loss and waste alone is FL usually takes place due to inefficiencies in the food supply chain
$49.5 billion in 2016, which is equal to 3% of the country’s gross such as inadequate management in storage facilities, technological
domestic production (Gooch and Nikkel, 2019). Due to the extent of failures in refrigeration, and poor infrastructure during trans-
food that is being wasted annually across the world, and the asso- portation (FAO, 2014a). An important part of food loss is ‘food waste
ciated enormous environmental and socio-economic burden of this (FW)’, which refers to the intentional removal of food from the FSC,
food waste (FW), authorities are prioritizing a progressive reduction that is still fit for consumption. FW is more often considered to
of FW generation (Abdulla et al., 2013). Quantifying the amounts of occur by choice, for example, surplus preparation and neglecting
FW generated and analyzing its composition are considered crucial which results in spoiled, expired or surplus uneaten food (CEC,
steps in reducing FW generation at each stage of the food supply 2019). However, some studies distinguish FL and FW solely based
chain (Xue et al., 2017). A major obstacle to quantify FW, especially at on the stage in which it is generated (van der Werf and Gilliland,
the household level, is the lack of standard methodologies, which 2017). These studies consider the decrease in edible food mass at
has resulted in utilization of numerous methodologies that are the production, postharvest, and processing stages as FL, and food
substantially different from each other (van Herpen et al., 2019). lost at the level of retailers and consumers as FW (van der Werf and
Gilliland, 2017). This common distinction between FL and FW is
1.1. Aims of current study illustrated in Fig. 1.
FW is further classified into a number of categories, edible and
Considering the absence of a critical and comprehensive review inedible avoidable, possibly avoidable and unavoidable (Quested
of FW quantification methods, the aim of the current study is to and Johnson, 2009). The latter categorization was first introduced
compare household FW quantification methods, with a focus on by the Waste Reduction Action Program (WRAP) of the United
methods addressing composition analysis. Recent studies related to Kingdom in 2009 (Quested and Johnson, 2009) and has been
FW quantification at the household level were reviewed system- adapted more frequently in FW studies ever since (e.g. Delley and
atically to understand the strengths and limitations of the quanti- Brunner, 2017; Edjabou et al., 2016; Parizeau et al., 2015; van der
fication methods each study has used. This paper presents a critical Werf et al., 2018). Avoidable FW is the food that was edible at
comparison of the methods and indicates the strengths and limi- some point prior to disposal. For example, food that has spoiled or
tations of each method, which is useful for future researchers in reached its best before date and discarded are considered avoidable
selecting the best method that caters to their study requirements. (FUSIONS, 2014). Food that some individuals eat but others do not,
The first section of the paper presents background information such as potato peels, beet greens, and bread crusts is considered
and the study context related to food loss and waste and justifies the possibly avoidable. Food that is not edible under normal circum-
need of the present study. Section two presents the methodology stances, such as orange peels, meat bones, and egg shells is
used for the current literature review. In the results and discussion considered unavoidable (Quested and Johnson, 2009). However,
section, a short analysis of bibliographic information is presented the classification of FW into these categories may depend on cul-
first, followed by the in depth analysis of four different FW quanti- tural factors, and what food belongs to each category may change
fication methods: surveys and interviews; kitchen diaries; waste over time (Kummu et al., 2012). Although categorization based on
audits; and secondary data. Strengths and weaknesses of each of the avoidability or edibility are the most frequently used methods, a
methods are discussed in detail and an overview of all methodolo- range of other classification methods can be found in FW literature.
gies and their applicability are discussed. The recommendations and (Garcia-Garcia et al., 2017). presents 9 different FW classification
conclusions section presents guidance for future researchers on the methods based on different indicators such as origin (animal/plant
applicability of each method for different study contexts. based), complexity (single/mixed product), presence of animal
products, treatment (processed/unprocessed), packaging etc.
2. Background
2.2. Impacts of food loss and waste
2.1. Food loss and waste
Food loss and waste that occurs throughout the FSC has recently
FAO (2014a,b) defines ‘food’ as any substance which is intended attracted global attention due to its adverse impacts not only on the
for human consumption, either processed, partially processed, or environment, but also on the economy and society on a broader
raw. This definition is inclusive of any drinks, chewing gum, and scale (Garcha, 2017). When the economic burden is considered, the
any other substances that have been used during manufacturing, associated monetary value of lost food alone across the globe is
processing and treatment of ‘food’. However, the food that is approximately $750 billion annually (FAO, 2011). According to FAO
 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722 3

Fig. 1. Overview of food supply chain and the distinction between food loss and food waste (based on the information presented by Gooch and Nikkel (2019) and van der Werf and
Gilliland (2017)).

estimates, the total economic cost of food wastage is about $1 One of the key approaches towards handling FW issues would
trillion each year (FAO, 2014b). be measuring, tracking and reporting the quantities and composi-
Environmental impacts related to FW (associated with impacts tion of waste generated across the food supply chain, as it is not
of producing food that is wasted), range across a number of con- possible to manage something that is not measured (Rajan et al.,
cerns including use of water, land, fertilizer and energy, as well as 2018). However, several researchers have identified lack of
loss of biodiversity and climate change. According to FAO estimates, knowledge and methodological gaps in quantifying FW generation
the annual global food loss and waste has a carbon footprint of 3.3 as the major obstacles in addressing FW generation (Chaboud,
Giga tonnes of CO2 equivalence without accounting for GHG 2017; Edjabou et al., 2016; Eriksson et al., 2012; Parizeau et al.,
emissions from land use change, ranking FW the third highest GHG 2015). When household FW generation is considered, it is essen-
emitter after USA and China (FAO, 2013) In addition to economic tial to clearly understand the quantity and the composition of FW
and environmental costs, FW incurs a vast range of social costs. generated in order to change the household waste behavior
Resource depletion and pollution from agricultural production of (Parizeau et al., 2015). Without this knowledge, it is difficult to
lost or wasted food leads to food security risks, loss of livelihoods, target, prioritize, and design actions to prevent and reduce food
individual and societal health costs, and loss of well-being and loss and waste.
societal value due to loss of habitat and landscape amenities (FAO, The lack of a standard and widely accepted methodology for
2014b). quantifying amounts and composition of household FW has led
researchers to adopt numerous methods, including waste audits,
2.3. Household FW and the gaps in existing knowledge kitchen diaries, surveys, and estimations based on secondary data
(Bra€utigam et al., 2014). Many recent publications have recog-
Although food is lost or wasted throughout the food supply nized the inconsistencies in these numerous methodologies for
chain, many studies report that in developed countries, the highest quantifying FW as a limitation which restricts valid comparisons
percentage of food loss/waste happens in the post consumption among different studies (Edjabou et al., 2016; Elimelech et al.,
stage, especially at households (Br€ autigam et al., 2014; Gooch and 2018; Parizeau et al., 2015; van Herpen et al., 2019; Xue et al.,
Felfel, 2014; Parfitt et al., 2010). Although, a recent study in Can- 2017).
ada reported that households are responsible for only 14% of total The previous review articles by Xue et al. (2017) and van der
FLW and 21% of avoidable FW (Gooch and Nikkel, 2019), there is still Werf and Gilliland (2017) present an overview of FW research
a significant uncertainty regarding the quantity of FW generated at that was published before 2014 and 2015, respectively. Moreover,
households. Large quantities of FW from the household sector both these studies focused on the entire food supply chain, thus
result in high costs for collection and transport, as well as for overlooking the specific methodological barriers for quantifying
separation and treatment in waste management facilities household FW. Although Schanes et al. (2018) reviewed studies on
(Bra€utigam et al., 2014). Due to this significant contribution, household FW until the year 2017, their focus was specifically on
focusing on household FW is important as it plays a major role in FW practices as well as distilling factors that foster and impede the
meeting FW reduction targets at local as well as global level. For generation of FW at the household level. Given that none of the
instance, the Goal 12 of United Nations Sustainable Development publications have attempted to compare the strengths and weak-
Goals (SDGs); “Ensure sustainable consumption and production nesses of each methodology in a comprehensive manner, we pro-
patterns” includes amongst its targets to “halve per capita global vide a detailed review of household FW quantification methods to
food waste at the retail and consumer level, and reduce food losses identify gaps for future research.
along production and supply chains by 2030" (United Nations,
2015). Thus, it is evident that additional attention is needed to 3. Methodology
understand and reduce household FW due to its substantial
contribution to FW generation. In this study, a systematic literature search approach was used,
4 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722

employing a scholarly data base and pre-defined key words, as analyze composition’, ‘response rate’, ‘sample size’, ‘ability to track
shown in Fig. 2. Through initial screening of related literature, a set alternative means of disposal’, ‘ability to track root causes’ and
of keywords to represent FW scenarios, specifically at the house- ‘technical expertise.’ Regardless of the number of phrases
hold level, was selected. The selected key words were “food waste” mentioned in the study related to each concept, the researcher only
or “residual waste” or “household waste” or “food loss” or “food loss counted the presence of each concept as one occurrence, but noted
and waste” or “green bin” or “organic waste” and “households” or the context of the phrase to obtain a more nuanced analysis. For
“homes” or “consumer” or “residential”. The above keywords were example, if an article mentioned a method as being ‘expensive’, the
used to search Scopus database for publications and peer reviewed conditions that made it were noted, such as ‘expensive if subjects
journal articles related to household FW published within the time are compensated.’ Results for each paper were summarized and
frame of January 1, 2010 to April 15, 2019. then synthesized across all the papers into broader categories of
The initial screening for selecting the relevant studies was done strengths and weaknesses (e.g. Strengths: Large sample size is
by reading the titles of the studies. At this stage, in order to prevent possible; Weakness: Needs a significant effort from the participants
relevant studies from being excluded, all the studies that addressed such that tapering enthusiasm of respondents can be problematic).
any component related to FW/organic waste/residual waste were
selected. This resulted in 366 studies, which were then subjected to 4. Results
the second step of screening.
The second screening was done by reading the abstracts and The results are presented in two major categories: 1) meta-
reviewing the full articles where necessary to select studies that analysis of studies reviewed; and 2) critical review of available
specifically addressed quantification and/or composition analysis of methods for quantifying household FW.
household FW. The following inclusion/exclusion criteria were used
for the second screening: (a) published within January 1, 2010 and 4.1. Studies reviewed
April 15, 2019; (b) published in English language; and (c) quantified
and/or analyzed composition of FW generated at households. Out of 4.1.1. Analysis of bibliographic information
the 366 studies that were subjected to the second screening, 45 The bibliographic information of the 366 studies that passed the
studies were selected for the in-depth review of methodologies. initial screening of titles shows how the number of publications
While screening out the studies relevant to the above mentioned related to FW has increased steadily over the last decade, with a
criteria, all 366 articles were broadly categorized into 9 categories three-fold increase from 2013 to 2018 (Fig. 3). This justifies the need
to obtain a broad overview of the related literature. of the current review, which captures the more recent studies that
Strengths and weaknesses of each FW quantification method were not included in the previous reviews by van der Werf and
were identified by conducting a simplified conceptual content Gilliland (2017) or Xue et al. (2017).
analysis of the selected journal articles, using general guidelines The selected 366 studies from the first screening are grouped ac-
provided by Thomas and Harden (Thomas and Harden, 2008) and cording to country or territory to understand the geographical dis-
White and Marsh (White and Marsh, 2006). The research question tribution of the recent FW studies. This classification was done based
was “What are the strengths and weaknesses associated with on the country that the study was carried out or the country of the first
various FWquantification methods?” Using an inductive approach, author (in studies that did not have a clear geographical boundary).
each paper was read thoroughly to identify phrases that repre- According to this analysis, most of the FW research is concentrated in
sented various concepts related to strengths and weaknesses of industrialized countries, especially in Europe, where more than 250
quantification methods, such as: ‘accuracy’, ‘cost/expensive/inex- out of 366 studies were carried out (Fig. 4). The highest number of
pensive’, ‘subjective/objective/bias’, ‘composition analysis/ability to studies is recorded in the United Kingdom (64 studies) followed by 53
in the USA, 35 in Italy, 31 in Sweden and 26 in Denmark. This attention
to FW in Europe may be due to the numerous action plans, regulations
and legislations on FW put forward by the European Union (Vittuari
et al., 2015) during the past decade. In contrast, only a handful of
studies have been carried out in developing countries, especially in
South Asian or African region. This may be because in developing
countries food is seldom wasted after purchasing as poverty and

Fig. 3. Number of articles related to food waste published from 2010 to May 2019
within the search criteria of the current study (This does not represent all publications
related to food waste but only the ones that appeared within the search criteria of the
Fig. 2. Summary of the literature review methodology. current study).
 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722 5

Fig. 4. Number of articles by country from 2010 to 2019.

limited income make it unaffordable to waste food, as some re- shown as Category 3 in Table 1), this categorization provided a
searchers argue (Ericksen, 2008; FAO, IFAD & WFP, 2015). broad overview of the selected literature (Table 1).
The top five academic journals ranked in order of number of The second step of the screening process led to the categoriza-
publications were Waste Management (47 articles), Journal of tion presented in Table 1. This categorization was primarily based
Cleaner Production (38 articles), Resources Conservation and upon the main objective or the primary research question each
Recycling (27 articles), British Food Journal (17 articles), and Waste study was attempting to address. All studies in which the main
Management and Research (12 articles). objective was to assess the food wasting behaviors, and not to
quantify FW were included in Category 1. During this second
4.1.2. Categorization of articles screening, it was observed that the majority of household FW
As explained in the methodology section, all the articles that studies have focused on assessing behavioral aspects of FW gen-
passed the first screening were then categorized into nine broad eration, rather than attempting to quantify the actual amounts.
categories according to their study objectives. Although the main Most of these studies used surveys or interviews to study attitudes
intention of this second screening was to select the most relevant and behaviors shaping food wasting behaviors at households. Some
articles that quantified FW generated at households (which is of the studies specifically looked at parameters such as packaging,
6 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722

Table 1
Categorization of articles according to their study objectives.

Category Description # of articles

1 Studies on food wasting behaviors, drivers and barriers 116

2 Studies that quantified residual household waste including organics (where study objective is not food waste specifically) 54
3 Studies that quantified the amount and/or analyzed composition of household food waste 45
4 Studies with a broader scope that covered more than one stage of food supply chain 32
5 Studies that focus on other stages of consumer food waste (retail/food services) 17
6 Studies that discussed Climate Change aspects of food waste 14
7 Policy reviews and policy implications related to food waste 13
8 Literature reviews and Meta-Analysis 6
9 Other (studies that could not be categorized into any of the above categories) 67
TOTAL (*Two articles were not accessible) 364

shopping behavior, price and suboptimal food in relation to wasting

behaviors. Since these studies did not specifically look at quantities
or composition of FW at households, all studies in this category
were excluded from further review.
Category 2 includes the studies that quantified residual house-
hold waste, including organic wastes. All studies in this category
were focusing on all streams of residual household waste including
organics, recyclables and garbage. Although some of these studies
used similar methods discussed in this review such as surveys and
waste audits, those methods were used in a rather generic manner
and not specific for FW. For instance, in residual household waste
audits, composition analysis would focus on categorizing the waste
into categories such as paper, plastic, metal and organics; or re-
cyclables, organics and garbage. These studies did not further
analyze the organic fraction and thus, overlooked the specific pa-
rameters that are important in a FW composition analysis. As these Fig. 5. Number of articles based on the quantification method used for Category 3.
studies lacked the specificity for quantifying FW, this category was
also excluded from further review. However, there were 45 studies
survey) and primarily rely on a third party other than the re-
that specifically aimed at quantifying the amount and/or analyzing
searchers (van Herpen et al., 2019). They can also be researcher-
the composition of FW generated at households. Since the current
administered (researcher asks questions and records answers
review aims at identifying and analyzing the methods of FW
from the respondent via telephone or in-person). In surveys, re-
quantification and composition analysis, studies in this category
sponses are recorded based on a single instance. Measures of FW
were further reviewed for their methods.
typically include absolute or frequency measures (how often),
The key word search brought up a large number of studies that
visually-based measures (visual aid provided to judge quantity), to
could not be included in any of the above three categories.
proportional waste measures (as a percentage or fraction of food
Depending on their main focus, they were categorized into seven
consumed) (CEC, 2019).
more categories (Category 4 to 9) as shown in Table 1.
Of the 45 articles reviewed, six articles primarily used surveys as
the data collection method (Table 2). Six other studies that coupled
4.2. FW quantification and composition analysis methods
surveys with another method/methods will be discussed in section
4.2.5.
All 45 articles in Category 3 were thoroughly reviewed to
Only one study was researcher-administered at the visiting
identify their methodologies and to draw comparisons. FW quan-
households (Zhang et al., 2018). Out of the self-administered sur-
tification methodologies identified in the analysis were: (i) surveys
veys, one study (Visschers et al., 2016) sent out a pencil and pen
and questionnaires; (ii) kitchen diaries (self-reported); (iii) esti-
survey by mail, while the remaining four studies used online sur-
mations based on secondary/aggregate data; and (iv) waste audits
veys. All surveys included a section where respondents’ de-
(direct measurement of wet weight). There were a number of
mographic information was recorded. Aschemann-Witzel et al.
studies that incorporated more than one of the above methods
(2019) asked respondents to recall only their most recent FW
(Rispo et al., 2015; Khalid et al., 2019; Sosna et al., 2019; Xu et al.,
instance, while all other studies asked respondents to recall FW
2016), while there were also studies that compared two or more
generation over a week. Most of the studies also considered the
methods (Delley and Brunner, 2018; Giordano et al., 2018; van
type of food being wasted and the reason for food wastage.
Herpen et al., 2019). The number of articles reviewed in the cur-
Nevertheless, there is inconsistency on what is being recorded,
rent study for different quantification methods are presented in
particularly with respect to amount of waste, which was recorded
Fig. 5.
as relative to absolute values.

4.2.1. Surveys
In this review, surveys are defined as a method which utilizes a 4.2.1.1. Strengths and weaknesses of surveys. The strengths and
questionnaire as a tool to capture self-reported quantities and the weaknesses of surveys in quantifying FW as discussed in the
composition of FW generated within a respondent’s household, and reviewed studies are summarized in Table 3.
sometimes includes questions on demographics and behavior Surveys are considered a cost-effective method to generate
related to FW. Surveys can be self-administered (respondents rough quantitative estimates of FLW at every stage of the food
answer the questions either online or with a paper and pencil supply chain (CEC, 2019). Although some researchers argue that the
 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722 7

Table 2
Studies that used surveys to quantify household food waste.

Article Country Sample Size Method

Aschemann-Witzel et al. (2019)1 Uruguay 540 households Survey (online) with open-ended questions asking respondents to describe their most
recent food wasting incident in terms of what was wasted, why and when.
€ rissen et al. (2015)2
Jo Italy & Germany 857 respondents Survey (online) of employees of two scientific institutions.

Covered only avoidable FW, studied household behavior (shopping, food preparation
and eating) related to FW over a week.
3
Lanfranchi et al. (2016) Italy 500 respondents Survey (online) estimated the food waste at five levels; “much, relatively, a little, very
little, nothing”
Martindale (2014)4 UK 100 respondents Survey (online) estimated the proportion of frozen and fresh food wasted daily using
graphical illustrations. These estimates were used to calculate a waste index.
Visschers et al. (2016)5 Switzerland 796 households Survey (paper and pencil mailed to households). Self-reported FW indicated the
frequency of wasting and the amount across 11 different food categories over a week.
6
Zhang et al. (2018) China 418 households Explored the nature of the FW produced, including its quantity as a percentage of the
food consumed, the proportion of avoidable and non-avoidable waste and the
composition of the FW generated

response rate for surveys could be rather low unless a monetary participants considering that wasting food is an unintentional
incentive is provided for respondents (van Herpen et al., 2019), the behavior. It would be difficult to recall exact FW quantities over a
ever-increasing access to internet, administering online surveys given time period (Aschemann-Witzel et al., 2019).
allow researchers to reach more people, thus resulting in a large Secondly, self-estimation without actual weighing can be
sample size (e.g., 796 households in Visschers et al. (2016), and 857 inaccurate. There are some approaches used to mitigate this.
households in Jo €rissen et al. (2015)). Martindale (2014) used illustrations to help respondents decide
Importantly, surveys can be quite beneficial when the the amount of FW (shapes proportional in size to the amount of
researcher does not have direct access to FW disposed by house- FW), while Visschers et al. (2016) used portion size (e.g. x
holds (Martindale, 2014). Surveys may also capture FW disposed number of portions; portion size being one handful of food
using alternative means such as composting, pet food, and over the served) as a measure of FW as well as frequency of occurrence
sink (Aschemann-Witzel et al., 2019). In addition, surveys often (e.g. number of times per week/month, to less often or never) in
collect demographic information of the participants, making it each type of FW category. Zhang et al. (2018) asked respondents
possible to examine correlations between amount of waste gener- to recall the amount of FW as a percentage of total household
ation and other factors such as income, age, number of family waste as well as the amount of avoidable FW as a percentage of
members etc. total FW. Aschemann-Witzel et al. (2019) asked the respondents
A notable strength in surveys is their ability to examine FW to record their most recent food wasting incident instead of
composition, as well as root causes, drivers, and barriers of recording the daily or weekly FW amounts, arguing that it is
food wasting behaviors. For example, Aschemann-Witzel et al. more accurate than arbitrary estimations in overall terms.
(2019) studied the categories of food being wasted (e.g. rice, However, this may not be representative of the FW quantities
vegetable, baked goods, etc.), the reason for wasting (e.g. bad over a period of time considering that the study takes into ac-
quality, prolonged storage, reached expiration date, etc.), and count only a single instance for each participant. A study by
the situation in which waste occurred (e.g. food preparation, Lanfranchi et al. (2016) asked the respondents to express their
eating, cleaning fridge, etc.). Visschers et al. (2016) recorded FW amounts in terms of categories ranging from ‘much’, ‘rela-
the frequency of FW occurrence in 11 different food groups, tively’, ‘a little’, ‘very little’, to ‘nothing’. However, this is highly
while also assessing intentions to avoid FW, attitudes and subjective because an individual’s perception of quantity in terms
behavioral norms. of ‘much’ or ‘a little’ could differ from that of another individual
Despite these strengths, surveys have a range of limitations. One which would result in unreliable estimates. Regardless of the
of the major limitations in using surveys for FW quantification is applicability and representativeness of each of these methods,
that surveys solely depend on recall and self-reported estimates the inconsistency of measurement leaves little to no room for
and measurements, which leads to several problems. comparison across different studies. Comparison of FW quanti-
Firstly, the measures are inconsistent across different studies ties and composition among different studies is important to
making it challenging to draw comparisons among them. For understand how FW generation changes between different
instance, some studies require the respondents to record their FW geographic regions depending on their socio-cultural, economic
in absolute measures of weight such as in grams per meal or per or political background, or in the same geographic region over
day or per week (Jo € rissen et al., 2015). This could be difficult for time to develop appropriate reduction strategies.

Table 3
Strengths and weaknesses of surveys.

Strengths Weaknesses

 Possible to assess FW composition1,2,3,4,5,6  Self-reported FW quantities can be less accurate, with a tendency to
 Can obtain data on demographic and other characteristics of respondents1,2,4,5,6 underestimate1,2,4,5,6
 Wide reach and allows larger sample size1,2,3,5,6  Response rate can be low2,3,4,5
 Can assess causes of FW1,2,3,5,6  Respondents may be inclined to give socially-desirable responses1,2,5,6
 Relatively low cost1,5Can assess the effectiveness of interventions related to  Researcher-administered surveys can be costly and time consuming1
behavioral aspects2
 Researcher-administered surveys allow clarification of questions5
8 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722

Thirdly, respondents might intentionally under-report actual condition of food at the point of disposal (whether food is spoiled or
amounts of FW, or unintentionally over/under-estimate the reached best before date) as well as the means of disposal (garbage/
amount as wasting food is considered a negative and non- composting/over the sink) which captures all streams of waste
appealing behavior in society; this would lead to incorrect esti- regardless of whether they end up in garbage bin or not
mations of actual FW generation (Aschemann-Witzel et al., 2019). (Silvennoinen et al., 2014).
Further, Hebrok and Boks, (2017) argue that since self-reporting Finally, kitchen diaries allow respondents to record the
depends on respondents’ ability to understand, recall and record, composition of FW in a more detailed manner than with surveys, as
this can affect the validity and accuracy of the results. Although pre- per research requirements. This method is capable of capturing not
announcing a survey may reduce this effect, it does not necessarily only the amount of food wasted, but also the type of food and
eliminate it since respondents may be inclined to give socially- whether it is avoidable or unavoidable (van Herpen and van der
desirable answers regardless (van Herpen et al., 2019). Thus, it is Lans, 2019). This is a major strength in FW research as it allows
evident that the accuracy of the estimates obtained from surveys the researcher to capture as much information as possible during
could be highly uncertain. the food wasting incident itself. Many researchers argue that di-
aries are a substitute for detailed observations in terms of observing
4.2.2. Kitchen diaries everyday situations, but without the effect of researcher being
In this review, kitchen diaries refer to the practice where the present. Thus, diaries can provide detailed information about the
respondents (typically residents of households) keep records of the daily habits where researchers would normally have no access, and
amount and nature of FW generated at their homes on a regular which otherwise would have been neglected (Richter and
basis. This method often requires the respondents to measure and Bokelmann, 2017). Diaries are also suitable to examine subjective
record the amount (i.e., weight or volume) and the type of food experiences, cognitions, and behavioral aspects of the participants,
being wasted, along with the reasons and situations under which especially to show how and why people act the way they do.
FW occurred (CEC, 2019). Unlike surveys, kitchen diaries record Although surveys are also capable of recalling information, the
amount of FW as it occurs, and usually over a longer period of time. strength of kitchen diaries over surveys is that diaries record data
Of the studies reviewed, five studies used kitchen diaries to repeatedly over a definite period of time, during the time of action.
quantify FW at households (Table 4). Sample size varied from 20 In that sense, kitchen diaries would be an ideal method to study
households to 385 households and the majority of the studies root causes of FW, what makes people waste food and what they
recorded FW over a period of one week. However, one study had a feel about it as opposed to studying how much waste is actually
duration of two weeks (Silvennoinen et al., 2014), while another generated.
recorded weekly FW amount for over a period of eight months One of the most significant limitations of using kitchen diaries
(AlMaliky and AlKhayat, 2012). for FW quantification at households is that it requires a great deal of
Two of the studies (Giordano et al., 2019; Richter and time and effort from both participant and the researcher (Richter
Bokelmann, 2017) focused on shopping habits and purchases to and Bokelmann, 2017). On the one hand, if the participants are
understand relationships between what was purchased and what required to weigh all their FW upon disposal (eg: AlMaliky and
was wasted. Only two studies evaluated FW composition (Giordano AlKhayat (2012)), it results in a high reporting burden on the
et al., 2019; Williams et al., 2012). respondent that could result in low participation rates as well as
tapering of enthusiasm of participants with time. This would often
4.2.2.1. Strengths and limitations of kitchen diaries. A summary of result in small sample size compared to the population of interest.
the strengths and limitations of kitchen diaries is provided in For instance, in the study by Silvennoinen et al. (2014), out of 3000
Table 5. invitations to participate in the study, only 700 people volunteered
A prominent strength of kitchen diaries over surveys is that self- resulting in a response rate of 23.3%. Out of the 420 households
reported amounts are likely to be more accurate and less uncertain selected according to the study criteria, 40 households did not
with kitchen diaries, because they do not rely on recall, instead finish the study acceptably. On the other hand, if the participants
asking respondents to record FW quantities at the time of disposal are allowed to measure and record their FW according to their own
(Richter and Bokelmann, 2017). This eliminates the possibility of convenience, such as using a variety of measurements (weight/
unintentional incorrect estimations, although social desirability volume/proportion), it would result in a significant burden on the
and awareness of the study objectives can still lead to behavioral researcher having to transform all measurements into a single
change or intentional under-reporting (Langley et al., 2010; van standardized unit (Richter and Bokelmann, 2017).
Herpen et al., 2019). Another frequently cited weakness related to using the kitchen
Another strength is that kitchen diaries often require re- diary method is that it could lead to behavioral changes in partic-
spondents to record the background information associated with ipants, which will limit the ability to capture ‘business-as-usual’
wasting of food, such as the reason for throwing away, the scenario. Since respondents would be more conscious about their

Table 4
Studies that used Kitchen Diaries to quantify FW at households.

Article Country Sample Size Approach

AlMaliky and AlKhayat (2012)7 Iraq 20 households Weighed and recorded weekly food purchases and waste for eight months using kitchen scales.
Giordano et al. (2019)8 Italy 385 households Shopping habits and FW quantities were recorded in a daily diary to explore the relationship
between purchasing discounted food and FW. Edible and inedible FW quantities recorded with
the type of food.
Richter and Bokelmann (2017)9 Germany 25 households Households kept a diary over a period of one week to document food purchases, storage, and
amounts wasted. No composition analysis.
Silvennoinen et al. (2014)10 Finland 380 households Participants weighed and recorded all FW upon disposal, along with the reason for disposing;
this was done over a period of two weeks.
Williams et al. (2012)11 Sweden 61 households Households measured (weight or volume) and recorded their FW over a period of one week
with reasons for disposing. FW was categorized into seven categories of food.
 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722 9

Table 5
Strengths and weaknesses of kitchen diaries method.

Strengths Weaknesses

Provides information about the type of food as well as the root  Possibility of intentional under-reporting8,9,10,11
causes8,9,10,11 Needs a significant effort from the participants such that tapering enthusiasm of respondents can be
Can provide descriptive information that could not problematic7,9,11
be captured by other methods8,9,10,11 Can be costly, especially if the participants are compensated11
Captures FW that does not go into a waste bin9,10,11 Method itself can lead to changes in behavior8
 High accuracy, especially compared to surveys7,11

wasting habits during the period of participation, recording FW participating households (Lebersorger and Schneider, 2011). Most
using kitchen diaries could act as a constant reminder and positive of the studies that employed this method have collected samples on
motivator for behavioral change (Langley et al., 2010), which could a weekly basis to be representative of the food habits of the
result in declining validity of results over the time period. household (Parizeau et al., 2015), while some studies repeated the
weekly measurements for a longer time period to avoid any bias
4.2.3. Direct measurement through waste audits (Stejskal et al., 2017). However, the main drawback of the weekly
The current review defines a waste audit as a method where the sample collection is that it could be difficult to identify or categorize
researcher will directly collect, physically separate waste streams, certain food items due to decomposition over time. Moreover, this
and weigh and categorize each fraction of waste to generate ac- method does not capture the food that was discarded apart from
curate figures (van Herpen et al., 2019). The current review study the garbage bin (sink drain/home composting/animal feed)
identified 14 studies that collected data primarily based on waste (Parizeau et al., 2015). Additionally, depending on the time of year,
audits, among which two studies also had a self-reporting FW amounts and composition could change (Edjabou et al., 2018).
component. The selected studies, their sample size and the FW Providing kitchen caddies for disposing of FW at households is
quantities reported are shown in Table 6. considered as a more inclusive approach as participants can be
In waste audits, FW from households is collected, physically instructed to dispose all their FW in the caddies to avoid disposal by
separated, weighed and categorized. Often, waste audits primarily other means (Elimelech et al., 2018). Yet, mistakenly throwing food
focus on the waste that is put out for collection by the households, into the regular bin out of habit or the concerns over social desir-
but the approach for collecting and analyzing samples could be ability associated with FW might lead to underreporting of quan-
quite different from one study to another. This study identified tities in this approach too (van Herpen and van der Lans, 2019). This
three specific approaches for sample collection: obtaining samples in turn introduces bias into the study as participants are conscious
at the curb when households have put out their waste for collection about their food wasting habits resulting in unintentional behav-
(Parizeau et al., 2015); providing kitchen caddies for households to ioral changes (Urrutia et al., 2019). Another limiting factor can be
dispose their FW instead of disposing into the regular garbage bin the participation effort involved, which although is relatively low
(Zan et al., 2018); and obtaining samples from the municipal for the household participants, effort for researchers can be quite
collection centers (Oelofse et al., 2018). high due to the requirement of visits to individual homes
(Elimelech et al., 2018). Moreover, ethical concerns are a frequently
overlooked limitation in waste audits, since researchers have direct
4.2.3.1. Strengths and limitations of waste audits. The summary of
access to waste generated by the participating households during
the strengths and limitations of waste audits is presented in Table 7.
the audit and can witness alcohol consumption, drugs or erotic
Compared to other waste quantification and characterization
material.
approaches, the biggest strength of waste audits is that it has high
Both of the above approaches usually consider each household
validity and data quality, as it does not rely on self-reported amounts
as a single data point, and thus are able to correlate waste gener-
(van Herpen et al., 2019). Since the researcher is recording FW
ation with household characteristics if needed. Moreover, these
quantities instead of the participants, it eliminates the probability of
approaches can also observe the changes in food wasting behavior
intentional under-reporting (Parizeau et al., 2015).
over a period of time, for instance before and after interventions.
Another strength is waste audits are capable of analyzing FW
The third approach where the samples are obtained at the
composition, even though a detailed study would be time
municipal collection point instead of individual houses is a less
consuming and arduous. Moreover, waste audits generate high
exhaustive method, however, it is not able to capture information
quality data that could be used for comparison with other
associated with individual households (Oelofse et al., 2018). A
geographical regions or over time although it demands a consid-
positive point of this approach is that it does not require the
erable scientific knowledge and physical resources such as trans-
researcher to visit households and the households are also unaware
portation and sorting facilities.
of the study, which eliminates any bias due to household behavior.
However, this method requires significant expertise, time and
cost (Parizeau et al., 2015; van Herpen et al., 2019). Although the
basic approach to waste audits is physically collecting and 4.2.4. Estimates based on secondary data
measuring FW, each waste audit can significantly differ according Another method that was found in the FW literature is using
to the sample (each household vs. whole municipality), sample size secondary data from previous FW studies at the household level,
(no of households), sampling duration (daily vs weekly), and along with consumption or national or regional FLW estimates, to
method of collection (curbside/municipal collection point/kitchen generate FW estimates at the household level. This method is
caddies). significantly different from all other household FW quantification
The most frequently used approach for waste audits is collecting methods since there are no observations or recording of FW
samples at the curb, when the households have set out their quantities at the household level. This is an indirect FW quantifi-
garbage bins for collection by the municipal truck. If the households cation method that uses modelling, use of proxy data or use of
were not informed about the study objectives prior to sampling, literature data to estimate FW quantities (Caldeira et al., 2017). The
this method imposes no bias or behavior alterations in the present review found eight studies that used such secondary data,
10 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722

Table 6
Studies that quantified household FW using waste audits.

Source Country Sample FW/ FW/ Approach

Size capita/ household/
week week

Lebersorger and Austria 130 0.640 kg Samples were picked up during the regular waste collection day for weighing and sorting. Sorted
Schneider (2011)12 according to avoidability, life cycle stage and packaging.
Parizeau et al. (2015)13 Canada 222 4.2 kg Source separated organics weighed on two subsequent garbage collection days collected at
curbside and analyzed for composition
van der Werf et al. Canada 900 2.4 kg FW composition audit data from 9 municipalities-curbside weekly collection e data from 2012 to
(2018)14 2015
Simunek et al. (2015)15 Czech 17 1.01 kg Daily logs for food waste and food purchase, for a period of three weekdays and one weekend, this
Republic later estimated to a weekly amount, food was discarded into separate plastic bags provided, and
afterwards audited to compare with the logs
Sosna et al. (2019)16 Czech 9 0.312 kg Waste collected by municipality, sampled separately for each household, cost is also recorded
Republic e0.637 kg
17
Stejskal et al. (2017) Czech 18 53e58.5 kg FW categorized into eight fractions, separately weighed and analyzed weekly
Republic per year
Edjabou et al. (2016)18 Denmark 1474 3.51 kg One week of waste from apartments, bi-weekly collection from single family households, sorted
into six food waste fractions, which were then sorted further into detailed fractions and then
grouped into an additional 11 food categories
Edjabou et al. (2018)19 Denmark 101 9.6 ± 4.5 kg Samples collected by municipal waste collection for a full week in each of the three seasons, spring,
e autumn summer, and winter. No composition analysis

9.9 ± 5.1 kg

e summer
9.2 ± 5.2 kg -
winter
Elimelech et al. (2018)20 Israel 192 1.82 kg Daily collection at doorstep, same day sorting, for seven consecutive days, composition analysis
included
rrez-Barba and
Gutie Mexico 41 2.24 kg Waste samples collected pre and post awareness program e weekly collection
Ortega-Rubio
(2013)21
Hanssen et al. (2016)22 Norway 220 3.76 kg Four municipalities with no source segregation, collected weekly by municipality collection from
each household separately weighed and analyzed, each food category weighed separately
Khalid et al. (2019)23 Pakistan 51 0.426 kg Plastic bags were given to the households to keep their one day (24 h) FW. Separate bags for each
FW fraction which was weighed separately
Chakona & Shackleton South 200 0.42 kg Self-reported waste audit - FW recorded and measured at disposal (in cups, Tea spoons,
(2017)24 Africa Table Spoons etc.)
Oelofse et al. (2018)25 South 65366 0.585 kg Bulk sampling with randomized grab sampling, bulk sampling - weighing the municipal collection
Africa each week (weighing trucks), random sample of 100e200 g taken from each truck load and
analyzed for composition.
Bernstad, 201426 Sweden 680 2.0e2.5 kg Source separated organics from all households and 50% of residual waste bins analyzed for
composition, several categories of avoidable and unavoidable FW considered

Table 7
Strengths and weaknesses of waste audits.

Strengths Weaknesses
12,13,14,15,16,17,18,19,20,22,23,24,25,26
 Can provide detailed information about the composition  Cannot capture FW disposed in alternative means other than waste
 Avoid the bias due to social desirability unless respondents are given special bin12,13,14,18,20,25,26
bins12,13,14,15,16,18,19,20,21,25  Need direct access to FW13,14,16,19,20,23,26
 Allows to track progress over time12,14,17,19,21,25,26 State of degradation of FW material can challenge the accuracy of
 Can capture the “business-as-usual” scenario without changing respondents’ measures12,13,18,19,20,24,25
behavior15,13,19,22,25  Can be relatively expensive and time consuming12,13,14,20,25,26
 Might not be able to track root causes12,14,16,22,25,26
 Need technical expertise14,18,20,25,26
 Ethical sensitivity13,14,20

often coupled with a modelling based approach, to estimate and landfills) (e.g. Nahman et al., 2012); understanding regional
quantities of FW generated at households (Table 8). differences in avoidable and unavoidable FW to target in-
In general, many such studies (Buzby and Hyman, 2012; terventions; and do scenario modeling for future reductions of
Vanham et al., 2015a,b) refer to global level estimates by FAO or avoidable FW (e.g. De Laurentiis et al., 2018).
United States Department of Agriculture (USDA) database (Xue
et al., 2017) and then combine with previously published studies 4.2.4.1. Strengths and limitations of secondary data. A main strength
from the same region or similar regions, as well as modeling to of this approach is that it is cost effective and less time consuming
address questions such as: the magnitude of the problem (raising than more direct methods (Table 9). Secondi et al. (2015) and Song
awareness) and the economic costs to society of household FW, et al. (2018) used data from previous nation-wide surveys and
both direct (consumer) and indirect (municipal waste collection modeling to calculate approximate FW quantities generated at
 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722 11

Table 8
Studies that estimated household FW using secondary data.

Article Country Approach

Buzby and Hyman (2012)27 United States of America (USA) USDA data was used to generate aggregated values for total food loss in US
De Laurentiis et al. (2018)28 European Union (EU) Household level avoidable & unavoidable waste of fresh fruits and vegetables are estimated for
the EU28; unavoidable - considering the inedible fraction of the purchased amount, avoidable
-based on results from previous studies
Lusk and Ellison (2017)29 N/A Modelling based approach using household production model by Becker (1965)
Nahman et al. (2012)30 South Africa Aggregate data from previous studies used to generate average values for South Africa
Reynolds et al. (2014)31 Australia Data from three complimentary Australian studies incorporated with WRAP data. Modelling
based approach using a weighted average method in conjunction with a Monte-Carlo simulation
Secondi et al. (2015)32 EU Data from 2013 Flash Eurobarometer survey (n. 388) was coupled with study carried out by the
BIO Intelligence Service (Monier et al., 2011), and used a modelling based approach to estimate
FW for EU-27 countries
Song et al. (2018)33 China Data from China Health Nutrition Survey was modelled using Bayesian Belief Network system to
identify reduction scenarios
34
Vanham et al. (2015a), (2015b) EU Data from FAO Food Balance Sheets were used in a statistical model

households in their research, whereas Nahman et al. (2012) and 5. Discussion

Reynolds et al. (2014) aggregated data from several previous
studies to obtain their estimates. This method is also appealing and In this research, 45 studies using five different methods to quantify
applicable in instances where the access to primary data is limited and analyze the composition of FW at household level were reviewed.
or not available. Thirdly, this method usually covers a large sample Considering the strengths and limitations of each of the methods
size, often estimations for entire regions (e.g, EU) or countries discussed in the previous sections, it is reasonable to argue that there
(Nahman et al., 2012). is no ‘one best’ method for FW quantification at household level.
However, the data collection and analysis approach varies with However, selection of the most appropriate method should depend
each study and in turn so does the accuracy of resulting estimates on the research question each study is trying to answer and the level
(Caldeira et al., 2017). The data based on USDA loss adjusted FWand of access to resources. Using the findings of the current review, we
FAO regional estimates has been used repeatedly in FW research, present a simple decision tree to guide future researchers to select the
although its applicability is questionable especially in the context of most appropriate household FW quantification method (Fig. 6).
developing countries (Xue et al., 2017). Xue et al. (2017) argue that If the study objective is quantification rather than composition
household waste statistics reported by FAO (2011) did not have a analysis, most accurate results can be obtained through a weight-
single measured data point in Asia or Africa. Thus, an inherent based waste audit, given that the researcher has direct access and
limitation in this method is the use of several assumptions and resources to collect and measure FW. For instance, many studies
large approximations rather than precise quantification through attempt to quantify household FW at municipal or provincial levels
measurement, which would result in FW estimates at household to assist in the policy and decision-making processes related to
level with a significant margin of error. Further, the aggregation of organic waste management; i.e., to establish or manage central
results from different studies that have used different approaches composting facilities, to promote home composting, to increase
can also increase the uncertainty of findings. waste diversion etc. For such requirements, estimating the quantity
of FW as accurately as possible is more crucial than the exact
4.2.5. Mixed methods approach composition, and a simple weight based audit would be most
The current review identified three studies (Group A in Table 6) appropriate. A weight-based waste audit should focus on
that have compared two or more of the above methods and another measuring the weight of FW in a larger sample (higher number of
eight studies (Group B in Table 6) that combined two or more of the households) and might not need an in-depth composition analysis.
above methods to estimate FW at households. Studies in Group A In an instance where resources are limited and collecting FW from
used two or more methods individually to quantify FW and then households is impossible, a kitchen diary method could be used.
drew comparisons among the results, whereas studies in Group B Although surveys can also estimate the quantities, accuracy would
present results based on a combination of multiple methods. A brief be lower since respondents are reporting quantities based on recall.
overview of these studies are given in Table 10. For studies in which the research question is more biased to-
Using a mixed method approach can be beneficial since it can be wards composition than quantity, a composition-based waste audit
designed in a way that strengths of one method would compensate could be carried out. In a composition-based waste audit, a
for the weaknesses of the other, and to address data gaps from researcher could focus more on categorizing FW into as many types
certain methods. For instance, when waste audits alone cannot as required (avoidable/unavoidable or according to type of food) in
observe the root causes and behavioral aspects associated with a smaller sample of households, in contrast to weight-based audits.
food wasting behaviors, this challenge could be overcome by Nevertheless, if access to collecting waste is restricted, the kitchen
coupling a waste audit with a follow up survey (Khalid et al., 2019; diary method could be designed to capture as much compositional
Rispo et al., 2015; Sosna et al., 2019). data as required.

Table 9
Strengths and weaknesses of using secondary data as a method.

Strengths Weaknesses
27,28,2930,33
 Beneficial when primary data is inaccessible  Highly impossible to collect accurate composition data27,30,31,32,33,34
 Can cover a large sample size27,28,30,32,34  Accuracy depends on the approach28,29,31,32,33,34
 Low cost and less arduous30,31,33,34  Cannot study root causes or food wasting behaviors28,29,32,34
12 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722

Table 10
Overview of the studies that used more than one method.

Article Country Approach

Group A - Comparison of methods

Delley and Brunner (2018) Switzerland Postal survey results compared with extrapolated results from a national waste compositional analysis
report
Giordano et al. (2018) Italy Survey followed by diary for one week, waste audit followed.
van Herpen et al. (2019) The Netherlands Compared general surveys, diaries, photo coding, kitchen caddies, and pre-announced survey questions
regarding a specific time period. In an experiment, respondents were asked to assess their food waste using
some or all of these methods depending on condition.
Group B - Studies that used multiple methods
Khalid et al. (2019) Pakistan Waste audit (kitchen caddies) followed by a face to face interview
Koivupuro et al. (2012) Finland Kitchen diary coupled with a follow up survey
Parizeau et al. (2015) Canada Waste audit with a follow up survey
Rispo et al. (2015) UK Waste audit (bulk sampling at collection points) followed by a survey
Song et al. (2018) China Survey data coupled with existing Life Cycle Assessment data
Sosna et al. (2019) Czech Republic Waste audit combined with observations, informal interviews and semi standardized interviews
Urrutia et al. (2019) Canada FW measurement by researcher for quantification and participant observations and interview data to
understand material and visceral dimensions of household FW
Xu et al. (2016) China Weekly waste audits followed up with a post program interview

Fig. 6. Simple decision tree to select appropriate household FW quantification methods. Note that resources available (e.g. time, personnel) need to be factored in with respect to
how many samples are needed.

Similarly, studies that focus on life cycle aspects of food systems generate more accurate and representative data without inducing
may require quantity of FW as well as composition. The best behavior change, audits are resource and cost intensive and require
approach to capture both these parameters is to do a composition- intense planning. Kitchen diaries can also be used to assess both
based waste audit. Sample size should be decided based on the quantity and composition at the same time by asking respondents
required accuracy and available resources. Although waste audits to measure and record exact quantities and composition. However,
 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722 13

there is a probability that response rate could decrease over time 6. Conclusions
due to the significant effort needed from the respondents.
Due to the variability in study designs even within a single FW has become an environmental as well as a societal issue with
method, it is extremely difficult to draw a valid comparison among a high impact in both developed and developing countries that has
the quantities of household FW in two countries or regions. For important policy implications. With households being the largest
instance, from the selected 14 studies (for the current review) that contributor to FW, especially in developed countries, valid mea-
used waste audit as the primary method, eight studies reported FW surements or quantifications of FW at households are important as
as a per household value while six remaining studies reported it as they provide the opportunity to assess the nature of FW, to draw
a per capita value. As another example, while Parizeau et al. (2015) comparisons across time, countries/regions, and/or consumer
reported that the average organic waste production in Southern groups, to examine causes of FW generation, and to assess the
Ontario (Guelph) households to be 4.2 kg per capita per week, van effectiveness of interventions.
der Werf et al. (2018) reported that the Southern Ontario house- This comprehensive review of recent studies (2010 To 2019) that
holds generate about 2.4 kg of organic waste per week. While quantified household FW analyzed the strengths and weaknesses of
Parizeau et al. (2015) conducted a waste audit in 222 households for five different FW quantification methods and concluded that there
a period of two weeks, van der Werf et al. (2018) estimated the FW is no ‘one best’ method for FW quantification at household level.
using waste audit data from approximately 900 households in nine Given the need to customize quantification methods based on the
municipalities over the period of four years. This illustrates how a research question being asked, we provided a decision-tree to aid
particular study method and sampling approach can significantly researchers in choosing a method. However, the notable issue with
impact the quality and comparability of results. having such diverse array of methods is that the researchers are
Another important factor is the ability of each method to unable to compare the FW scenarios for two different geographical
analyze the composition of FW since it is highly important to regions at a higher accuracy. Even the studies that used the same
know what is wasted in order to develop strategies aimed at FW methods (e.g. surveys) generate results that are not quite compa-
reduction (van Herpen and van der Lans, 2019). Surveys and rable even with a similar study due to the vast differences in the
kitchen diaries can obtain detailed information about the types of protocols within the same methods (generating FW figures as
food wasted. However, with surveys, respondents have to recall percentage of consumption vs frequency). Thus, it is still important
their food wasting incidents, and it is possible that people might to develop standard protocols for each of these methods so that the
not remember every food item they wasted. With kitchen diaries, researchers will have the ability to compare and contrast their
if the respondents are asked to fill out the diaries at the time of findings with similar studies around the world. Nevertheless, it
waste generation, it is quite possible to capture more accurate should be noted that the present study has few limitations in its
information about the types of food wasted. Nevertheless, reluc- approach towards literature review, where considering only one
tance to record actual behavior can challenge the accuracy of the database “Scopus” is the major limitation. This review only looked
records. In contrast, the FW audits where the researcher could go at studies in which the main aim was to quantify and/or analyze
through the waste collected at households, are able to generate composition of FW at households, and did not consider the studies
more accurate findings regarding the composition of FW as well as that were aiming at studying the attitudes, beliefs and behaviors
the quantity of each waste fraction (Edjabou et al., 2018). Never- associated with FW or the studies that looked at all residual
theless, FW tends to decompose faster, especially in warm cli- household waste in general without special emphasis on FW.
mates, making it hard to identify each fraction while sorting
(Edjabou et al., 2018). This could be avoided by decreasing the Declaration of competing interest
time between waste generation and sorting (e.g., collecting waste
samples daily or twice a week instead of weekly). Novel methods The authors declare that they have no known competing
like photo coding and using fridge cameras can also generate financial interests or personal relationships that could have
comparatively accurate compositional data than surveys (van appeared to influence the work reported in this paper.
Herpen and van der Lans, 2019). Although coupling two or more
methods in a single study would be able to eliminate the weak- Acknowledgement and funding
nesses of a single method, there is still a possibility that it might
not be comparable with another study. This research was funded by the Social Sciences and Humanities
The most significant issue regarding FW composition analysis is Research Council of Canada (Grant Number: 435-2017-0544). Au-
the lack of an internationally accepted standard classification for thors would also like to thank the valuable feedback received from
types of FW. Although many studies follow the categorization the reviewers of the Journal of Cleaner Production.
introduced by WRAP by classifying FW into avoidable, possibly
avoidable and unavoidable categories, the actual sorting of FW References
during the study depends on the perspective of the researcher or
the participant. Due to cultural and societal perceptions, what Abdulla, M., Martin, R., Gooch, M., Jovel, E., 2013. The importance of quantifying
people perceive as avoidable could be different from region to re- food waste in Canada. J. Agric. Food Syst. Community Dev. 3 (2), 137e151.
https://doi.org/10.5304/jafscd.2013.032.018.
gion. Researchers generally define the avoidability upon local un- AlMaliky, S.B., AlKhayat, Z., 2012. Kitchen food waste inventory for residential areas
derstanding, making it fundamentally difficult for a meaningful in Baghdad city. Mod. Appl. Sci. 6 (8), p45. https://doi.org/10.5539/
comparison among studies in different geographical regions mas.v6n8p45.
Aschemann-Witzel, J., Gime nez, A., Ares, G., 2019. Household food waste in an
(Lebersorger and Schneider, 2011). Although defining a standard emerging country and the reasons why: consumer’s own accounts and how it
classification can enable meaningful comparison, it is also impor- differs for target groups. Resour. Conserv. Recycl. 145, 332e338. https://doi.org/
tant to pay attention to local perceptions since the wasting behavior 10.1016/j.resconrec.2019.03.001.
Becker, G.S., 1965. A theory of the allocation of time. Econ. J. 75 (299), 493e517.
depends on the self-perception. https://doi.org/10.2307/2228949. JSTOR.
Bernstad, A, 2014. Household food waste separation behavior and the importance of
14 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722

convenience. Waste Management 34 (7), 1317e1323. https://doi.org/10.1016/j. Hanssen, O.J., Syversen, F., Stø, E., 2016. Edible food waste from Norwegian
wasman.2014.03.013. householdsddetailed food waste composition analysis among households in
Br€ €rissen, J., Priefer, C., 2014. The extent of food waste generation
autigam, K.-R., Jo two different regions in Norway. Resour. Conserv. Recycl. 109, 146e154. https://
across EU-27: different calculation methods and the reliability of their results. doi.org/10.1016/j.resconrec.2016.03.010.
Waste Manag. Res. 32 (8), 683e694. https://doi.org/10.1177/ Hebrok, M, Boks, C, 2017. Household food waste: Drivers and potential intervention
0734242X14545374. points for design e An extensive review. Journal of Cleaner Production 151,
Buzby, J.C., Hyman, J., 2012. Total and per capita value of food loss in the United 380e392. https://doi.org/10.1016/j.jclepro.2017.03.069.
States. Food Pol. 37 (5), 561e570. https://doi.org/10.1016/j.foodpol.2012.06.002. €rissen, J., Priefer, C., Bra
Jo €utigam, K.-R., 2015. Food waste generation at household
Scopus. level: results of a survey among employees of two European research centers in
Caldeira, C., Sara, C., Serenella, S., 2017. Food Waste AccountingdMethodologies, Italy and Germany. Sustainability (Switzerland) 7 (3), 2695e2715. https://
Challenges and Opportunities. Luxembourg, Luxembourg, p. 45. EUR 28988 EN. doi.org/10.3390/su7032695. Scopus.
https://www.researchgate.net/profile/Serenella_Sala/publication/322315393_ Khalid, S., Naseer, A., Shahid, M., Shah, G.M., Ullah, M.I., Waqar, A., Abbas, T.,
Food_waste_accounting_-_Methodologies_challenges_and_opportunities/ Imran, M., Rehman, F., 2019. Assessment of nutritional loss with food waste and
links/5a53904d458515e7b72f04ad/Food-waste-accounting-Methodologies- factors governing this waste at household level in Pakistan. J. Clean. Prod. 206,
challenges-and-opportunities.pdf. 1015e1024. https://doi.org/10.1016/j.jclepro.2018.09.138. Scopus.
CEC, 2019. Why and How to Measure Food Loss and Waste:A Practical Guide. Koivupuro, H.-K., Hartikainen, H., Silvennoinen, K., Katajajuuri, J.-M., Heikintalo, N.,
Commission for Environmental Coorporation (CEC), p. 60. http://www3.cec. Reinikainen, A., Jalkanen, L., 2012. Influence of socio-demographical, behav-
org/islandora/en/item/11814-why-and-how-measure-food-loss-and-waste- ioural and attitudinal factors on the amount of avoidable food waste generated
practical-guide-en.pdf. in Finnish households. Int. J. Consum. Stud. 36 (2), 183e191. https://doi.org/
Chaboud, G., 2017. Assessing food losses and waste with a methodological frame- 10.1111/j.1470-6431.2011.01080.x. Scopus.
work: insights from a case study. Resour. Conserv. Recycl. 125, 188e197. https:// Kummu, M., de Moel, H., Porkka, M., Siebert, S., Varis, O., Ward, P.J., 2012. Lost food,
doi.org/10.1016/j.resconrec.2017.06.008. wasted resources: global food supply chain losses and their impacts on fresh-
De Laurentiis, V., Corrado, S., Sala, S., 2018. Quantifying household waste of fresh water, cropland, and fertiliser use. Sci. Total Environ. 438, 477e489. https://
fruit and vegetables in the EU. Waste Manag. 77, 238e251. https://doi.org/ doi.org/10.1016/j.scitotenv.2012.08.092.
10.1016/j.wasman.2018.04.001. Lanfranchi, M., Calabro , G., Pascale, A.D., Fazio, A., Giannetto, C., 2016. Household
Delley, M., Brunner, T., 2018. Household food waste quantification: comparison of two food waste and eating behavior: empirical survey. Br. Food J. https://doi.org/
methods. Br. Food J. 120, 1504e1515. https://doi.org/10.1108/BFJ-09-2017-0486. 10.1108/BFJ-01-2016-0001.
Delley, M., Brunner, T.A., 2017. Foodwaste within Swiss households: a segmentation Langley, J., Yoxall, A., Heppell, G., Rodriguez, E.M., Bradbury, S., Lewis, R.,
of the population and suggestions for preventive measures. Resour. Conserv. Luxmoore, J., Hodzic, A., Rowson, J., 2010. Food for Thought? d a UK pilot study
Recycl. 122 (Complete), 172e184. https://doi.org/10.1016/ testing a methodology for compositional domestic food waste analysis. Waste
j.resconrec.2017.02.008. Manag. Res. 28 (3), 220e227. https://doi.org/10.1177/0734242X08095348.
Edjabou, M.E., Boldrin, A., Astrup, T.F., 2018. Compositional analysis of seasonal Lebersorger, S., Schneider, F., 2011. Discussion on the methodology for determining
variation in Danish residual household waste. Resour. Conserv. Recycl. 130, food waste in household waste composition studies. Waste Manag. 31 (9e10),
70e79. https://doi.org/10.1016/j.resconrec.2017.11.013. 1924e1933. https://doi.org/10.1016/j.wasman.2011.05.023. Scopus.
Edjabou, M.E., Petersen, C., Scheutz, C., Astrup, T.F., 2016. Food waste from Danish Lusk, J.L., Ellison, B., 2017. A note on modelling household food waste behaviour.
households: generation and composition. Waste Manag. 52, 256e268. https:// Appl. Econ. Lett. 24 (16), 1199e1202. https://doi.org/10.1080/
doi.org/10.1016/j.wasman.2016.03.032. 13504851.2016.1265070.
Elimelech, E., Ayalon, O., Ert, E., 2018. What gets measured gets managed: a new Martindale, W., 2014. Using consumer surveys to determine food sustainability. Br.
method of measuring household food waste. Waste Manag. 76 (Complete), Food J. https://doi.org/10.1108/BFJ-09-2013-0242.
68e81. https://doi.org/10.1016/j.wasman.2018.03.031. Monier, V, Mudgal, V, Escalon, C, O’Connor, T, Gibon, G, Anderson, H, et al., 2011.
Ericksen, P.J., 2008. Conceptualizing food systems for global environmental change Preparatory Study on Food Waste across EU-27. Paris: BIO Intelligence Service.
research. Global Environ. Change 18 (1), 234e245. https://doi.org/10.1016/ Final report to the European Commission’s DG Environment. BIO Intelligence
j.gloenvcha.2007.09.002. Service., Paris.
Eriksson, M., Strid, I., Hansson, P.-A., 2012. Food losses in six Swedish retail stores: Nahman, A., de Lange, W., Oelofse, S., Godfrey, L., 2012. The costs of household food
wastage of fruit and vegetables in relation to quantities delivered. Resour. waste in South Africa. Waste Manag. 32 (11), 2147e2153. https://doi.org/
Conserv. Recycl. 68, 14e20. https://doi.org/10.1016/j.resconrec.2012.08.001. 10.1016/j.wasman.2012.04.012. Scopus.
FAO, 2011. Global Food Losses and Food Waste: Extent, Causes and Prevention. Food Oelofse, S., Muswema, A., Ramukhwatho, F., 2018. Household food waste disposal in
and Agriculture Organization of the United Nations. South Africa: a case study of Johannesburg and Ekurhuleni. South Afr. J. Sci. 114
FAO, 2013. Food Wastage Footprint: Impact on Natural Resources. Summary Report. (5/6) https://doi.org/10.17159/sajs.2018/20170284, 6e6.
Food and Agriculture Organization, p. 249. http://www.fao.org/3/ar429e/ Parfitt, J., Barthel, M., Macnaughton, S., 2010. Food waste within food supply chains:
ar429e.pdf. quantification and potential for change to 2050. Phil. Trans.: Biol. Sci. 365
FAO, 2014a. Definitional Framework of Food Loss (Global Initiative on Food Loss and (1554), 3065e3081 (JSTOR).
Waste Reduction) [Working Paper]. http://www.fao.org/fileadmin/user_upload/ Parizeau, K., von Massow, M., Martin, R., 2015. Household-level dynamics of food
save-food/PDF/FLW_Definition_and_Scope_2014.pdf. waste production and related beliefs, attitudes, and behaviours in Guelph,
FAO, 2014b. Food wastage footprint: full-cost accounting: final report. Food and Ontario. Waste Manag. 35, 207e217. https://doi.org/10.1016/
Agriculture Organization of the United Nations. http://www.fao.org/3/a-i3991e. j.wasman.2014.09.019.
pdf. Quested, T., Johnson, H., 2009. Household Food and Drink Waste in the UK: Final
FAO, IFAD, WFP, 2015. The State of Food Insecurity in the World. Meeting the 2015 Report. Wastes & Resources Action Programme (WRAP). http://www.wrap.org.
International Hunger Targets: Taking Stock of Uneven Progress. FAO. http:// uk/downloads/Household_food_and_drink_waste_in_the_UK_-_report.
www.fao.org/3/a-i4646e.pdf. 8438d59b.8048.pdf.
FUSIONS, 2014. FUSIONS Definitional Framework for Food Waste 2014.Pdf. Euro- Rajan, J., Fredeen, A.L., Booth, A.L., Watson, M., 2018. Measuring food waste and
pean Commission. https://www.eu-fusions.org/phocadownload/Publications/ creating diversion opportunities at Canada’s Green University TM. J. Hunger
FUSIONS%20Definitional%20Framework%20for%20Food%20Waste%202014.pdf. Environ. Nutr. 13 (4), 573e586. https://doi.org/10.1080/19320248.2017.1374900.
Garcha, N., 2017. Food Loss and Waste Solutions: Innovative Technologies and Best Reynolds, C., Mavrakis, V., Davison, S., Høj, S., Vlaholias, E., Sharp, A., Thompson, K.,
Practices, vol. 20. Ward, P., Coveney, J., Piantadosi, J., Boland, J., Dawson, D., 2014. Estimating
Garcia-Garcia, G., Woolley, E., Rahimifard, S., Colwill, J., White, R., Needham, L., 2017. informal household food waste in developed countries: the case of Australia,
A methodology for sustainable management of food waste. Waste and Biomass 32. Waste Management & Research. https://doi.org/10.1177/
Valorization 8 (6), 2209e2227. https://doi.org/10.1007/s12649-016-9720-0. 0734242X14549797.
Giordano, C., Alboni, F., Cicatiello, C., Falasconi, L., 2019. Do discounted food prod- Richter, B., Bokelmann, W., 2017. Explorative study about the analysis of storing,
ucts end up in the bin? An investigation into the link between deal-prone purchasing and wasting food by using household diaries. Resour. Conserv.
shopping behaviour and quantities of household food waste. Int. J. Consum. Recycl. 125, 181e187. https://doi.org/10.1016/j.resconrec.2017.06.006.
Stud. 43 (2), 199e209. https://doi.org/10.1111/ijcs.12499. Rispo, A., Williams, I.D., Shaw, P.J., 2015. Source segregation and food waste pre-
Giordano, C., Piras, S., Boschini, M., Falasconi, L., 2018. Are questionnaires a reliable vention activities in high-density households in a deprived urban area. Waste
method to measure food waste? A pilot study on Italian households. Br. Food J. Manag. 44, 15e27. https://doi.org/10.1016/j.wasman.2015.04.010.
https://doi.org/10.1108/BFJ-02-2018-0081. Schanes, K., Dobernig, K., Go €zet, B., 2018. Food waste mattersda systematic review
Godfray, H.C.J., Beddington, J.R., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F., of household food waste practices and their policy implications. J. Clean. Prod.
Pretty, J., Robinson, S., Thomas, S.M., Toulmin, C., 2010. Food security: the 182 (Complete), 978e991. https://doi.org/10.1016/j.jclepro.2018.02.030.
challenge of feeding 9 billion people. Science 327 (5967), 812e818. https:// Secondi, L., Principato, L., Laureti, T., 2015. Household food waste behaviour in EU-
doi.org/10.1126/science.1185383. 27 countries: a multilevel analysis. Food Pol. 56 (Complete), 25e40. https://
Gooch, M., Nikkel, L., 2019. The Avoidable Crisis of Food Waste: Technical Report. doi.org/10.1016/j.foodpol.2015.07.007.
The Second Harvest Canada and Value Chain Management International, p. 122. Silvennoinen, K., Katajajuuri, J.-M., Hartikainen, H., Heikkil€ a, L., Reinikainen, A.,
Gooch, M.V., Felfel, A., 2014. “‘27 Billion’” Revisited: the Cost of Canada’s Annual 2014. Food waste volume and composition in Finnish households. Br. Food J. 116
Food Waste. Value Chain Management Centre. (6), 1058e1068. https://doi.org/10.1108/BFJ-12-2012-0311.
Gutierrez-Barba, B.E., Ortega-Rubio, A., 2013. Household food-waste production and Simunek, J., Derflerova  Brazdova , Z., Vitu, K., 2015. Food Wasting: A Study Among
a proposal for its minimization in Mexico. Life Sci. J. 10 (3), 1772e1783. Scopus. Central European Four-Member Families, vol. 22, pp. 2679e2683.
 S.V. Withanage et al. / Journal of Cleaner Production 279 (2021) 123722 15

Song, G., Semakula, H.M., Fullana-i-Palmer, P., 2018. Chinese household food waste Vanham, D., Bouraoui, F., Leip, A., Grizzetti, B., Bidoglio, G., 2015a. Lost water and
and its’ climatic burden driven by urbanization: a Bayesian Belief Network nitrogen resources due to EU consumer food waste. Environ. Res. Lett. 10 (8)
modelling for reduction possibilities in the context of global efforts. J. Clean. https://doi.org/10.1088/1748-9326/10/8/084008. Scopus.
Prod. 202 (Complete), 916e924. https://doi.org/10.1016/j.jclepro.2018.08.233. Vanham, D., Bouraoui, F., Leip, A., Grizzetti, B., Bidoglio, G., 2015b. Lost water and
Sosna, D., Brunclíkova, L., Galeta, P., 2019. Rescuing things: food waste in the rural nitrogen resources due to EU consumer food waste. Environ. Res. Lett. 10 (8),
environment in the Czech Republic. J. Clean. Prod. 214, 319e330. https:// 084008 https://doi.org/10.1088/1748-9326/10/8/084008.
doi.org/10.1016/j.jclepro.2018.12.214. Scopus. Visschers, V.H.M., Wickli, N., Siegrist, M., 2016. Sorting out food waste behaviour: a
Stejskal, B., Malsova, A., Ba rekova , A., 2017. Comparison of family house and survey on the motivators and barriers of self-reported amounts of food waste in
apartment households bio-waste production and composition. Waste Forum households. J. Environ. Psychol. 45, 66e78. https://doi.org/10.1016/
237e243. j.jenvp.2015.11.007. Scopus.
Thomas, J., Harden, A., 2008. Methods for the thematic synthesis of qualitative Vittuari, M., Politano, A., Gaiani, S., Canali, M., 2015. Review of EU Legislation and
research in systematic reviews. BMC Med. Res. Methodol. 8 (1), 45. https:// Policies with Implications on Food waste.Pdf. FUSIONS, The European Com-
doi.org/10.1186/1471-2288-8-45. mission, ISBN 978-94-6257-525-7.
United Nations, 2015. Transforming Our World: the 2030 Agenda for Sustainable White, M., Marsh, E., 2006. Content analysis: a flexible methodology. Libr. Trends
Development. UN Publishing. https://sustainabledevelopment.un.org/content/ 55. https://doi.org/10.1353/lib.2006.0053.
documents/21252030%20Agenda%20for%20Sustainable%20Development% Williams, H., Wikstro €m, F., Otterbring, T., Lo
€ fgren, M., Gustafsson, A., 2012. Reasons
20web.pdf. for household food waste with special attention to packaging. J. Clean. Prod. 24,
Urrutia, I., Dias, G.M., Clapp, J., 2019. Material and visceral engagements with 141e148. https://doi.org/10.1016/j.jclepro.2011.11.044.
household food waste: towards opportunities for policy interventions. Resour. Xu, D.Y., Lin, Z.Y., Gordon, M.P.R., Robinson, N.K.L., Harder, M.K., 2016. Perceived key
Conserv. Recycl. 150, 104435. https://doi.org/10.1016/j.resconrec.2019.104435. elements of a successful residential food waste sorting program in urban
van der Werf, P., Gilliland, J.A., 2017. A systematic review of food losses and food apartments: stakeholder views. J. Clean. Prod. 134, 362e370. https://doi.org/
waste generation in developed countries. Proceedings of the Institution of Civil 10.1016/j.jclepro.2015.12.107.
Engineers - Waste and Resource Management 170 (2), 66e77. https://doi.org/ Xue, L., Liu, G., Parfitt, J., Liu, X., Van Herpen, E., Stenmarck, Å., O’Connor, C.,
10.1680/jwarm.16.00026. €
Ostergren, K., Cheng, S., 2017. Missing food, missing data? A critical review of
van der Werf, P., Seabrook, J.A., Gilliland, J.A., 2018. The quantity of food waste in the global food losses and food waste data. Environ. Sci. Technol. 51 (12),
garbage stream of southern Ontario, Canada households. PloS One 13 (6). 6618e6633. https://doi.org/10.1021/acs.est.7b00401.
https://doi.org/10.1371/journal.pone.0198470. Zan, F., Dai, J., Hong, Y., Wong, M., Jiang, F., Chen, G., 2018. The characteristics of
van Herpen, E., van der Lans, I., 2019. A picture says it all? The validity of photo- household food waste in Hong Kong and their implications for sewage quality
graph coding to assess household food waste. Food Qual. Prefer. 75, 71e77. and energy recovery. Waste Manag. 74, 63e73. https://doi.org/10.1016/
https://doi.org/10.1016/j.foodqual.2019.02.006. j.wasman.2017.11.051.
van Herpen, E., van der Lans, I.A., Holthuysen, N., Nijenhuis-de Vries, M., Zhang, H., Duan, H., Andric, J.M., Song, M., Yang, B., 2018. Characterization of
Quested, T.E., 2019. Comparing wasted apples and oranges: an assessment of household food waste and strategies for its reduction: a Shenzhen City case
methods to measure household food waste. Waste Manag. 88, 71e84. https:// study. Waste Manag. 78, 426e433. https://doi.org/10.1016/
doi.org/10.1016/j.wasman.2019.03.013. j.wasman.2018.06.010.