Waste Management xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Waste Management
journal homepage: www.elsevier.com/locate/wasman

The weight of unfinished plate: A survey based characterization of

restaurant food waste in Chinese cities
Ling-en Wang a, Gang Liu a,b,c,⇑, Xiaojie Liu a, Yao Liu a,d, Jun Gao e, Bin Zhou f, Si Gao g, Shengkui Cheng a
a
Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, 100101 Beijing, China
b
SDU Life Cycle Engineering, Department of Chemical Engineering, Biotechnology, and Environmental Technology, University of Southern Denmark (SDU), 5230 Odense, Denmark
c
Smart Cities Research Institute, School of Civil Engineering, Shenzhen University, 518060 Shenzhen, China
d
University of Chinese Academy of Sciences, 100049 Beijing, China
e
Waikato Management School, Waikato University, 3216 Hamilton, New Zealand
f
Department of Tourism, Ningbo University, 315211 Ningbo, China
g
School of Economics and Management, Hebei University of Technology, 300130 Tianjin, China

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

Article history: Consumer food waste has attracted increasing public, academic, and political attention in recent years,
Received 5 September 2016 due to its adverse resource, environmental, and socioeconomic impacts. The scales and patterns of con-
Revised 23 February 2017 sumer food waste, especially in developing countries, however, remain poorly understood, which may
Accepted 3 April 2017
hinder the global effort of reducing food waste. In this study, based on a direct weighing method and
Available online xxxx
a survey of 3557 tables in 195 restaurants in four case cities, we investigated the amount and patterns
of restaurant food waste in China in 2015. Food waste per capita per meal in the four cities was 93 g, con-
Keywords:
sisting mainly of vegetables (29%), rice (14%), aquatic products (11%), wheat (10%), and pork (8%). This
Food waste
Restaurant
equals to approximately 11 kg/cap/year and is not far from that of western countries, although per capita
Plate waste GDP of China is still much lower. We found also that food waste per capita per meal varies considerably
Consumer behavior by cities (Chengdu and Lhasa higher than Shanghai and Beijing), consumer groups (tourists higher than
Chinese cities local residents), restaurant categories (more waste in larger restaurants), and purposes of meals (friends
gathering and business banquet higher than working meal and private dining). Our pilot study provides a
first, to our best knowledge, empirically determined scales and patterns of restaurant food waste in
Chinese cities, and could help set targeted interventions and benchmark national food waste reduction
targets.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Food waste has gained increasing public, academic, and political
attention in recent years (Godfray et al., 2010; Gustavsson et al.,
2011; Charlebois et al., 2015; Gruber et al., 2015; Aschemann-
Witzel et al., 2015). For example, the United Nations’ recently
released Sustainable Development Goals has set a specific target
(SDG 12.3) for halving per capita global food waste by 2050. It
has been estimated that one third of global food production is lost
or wasted during the various phases of the food supply chain from
farm to table (Steinfeld et al., 2006; Kummu et al., 2012). Consumer
food waste at the consumption stage is usually high in developed
countries (Gustavsson et al., 2011; Kummu et al., 2012). In Europe,
⇑ Corresponding author at: SDU Life Cycle Engineering, Department of Chemical
Engineering, Biotechnology, and Environmental Technology, University of Southern
Denmark (SDU), 5230 Odense, Denmark.
E-mail addresses: gli@kbm.sdu.dk, geoliugang@gmail.com (G. Liu).
for example, as much as 42% of all food produced is wasted during
the final consumption stage (Monier et al., 2011).The U.S. wastes
31–40% of its post-harvest food supply, with a substantial portion
occurring at the consumer level (Neff et al., 2015). Such a high scale
of food waste implies that significant resource, energy, environ-
mental, and socioeconomic costs embodied in agrifood products
during production, transportation, processing, and retailing are in
vain (Koivupuro et al.,2010; Gustavsson et al., 2011; Lo and
Jacobson,2011; FAO, 2012; Eshel et al.,2014).
Addressing these consumer food waste issues requires an
understanding of their quantity and patterns in different consumer
segments, both in household and out of home. Examination of the
existing literature indicates that mass feeding away from home
often results in greater food waste than food preparation and con-
sumption within home (Youngs et al., 1983). However, most of
previous studies on food waste quantification focus more on
households than the hospitality sector. In addition, the investiga-
tions that have been carried out to examine the extent of food

http://dx.doi.org/10.1016/j.wasman.2017.04.007
0956-053X/Ó 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
2 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx
waste at out-of-home eating locations in the past decades (Grey
and Dubois, 1947; Banks and Collison,1981; Youngs et al.,1982;
Buchner et al., 2012; Silvennoinen et al.,2012; Parry et al., 2015)
covered largely only the western countries such as UK and US,
and this information remains poor for developing countries such
as China.
It is usually believed that in developing countries food loss
occurs mainly during the earlier stages of the food supply chain,
i.e., production and postharvest handing, due to lack of financial,
technical, and managerial resources, while food waste at the con-
sumer stage is assumed to be low, especially when compared to
that of developed countries (Parfitt et al., 2010; Dorward, 2012;
Kummu et al., 2012). However, this pattern may change as the
developing countries continue to urbanize, develop their economy,
and change their dietary structure (Parfitt, 2013; Thi et al., 2015).
For example, a study based on the scaling-up of household waste
streams in South Africa generated significantly higher amount of
household food waste than those widely cited but usually outdated
literature data (Nahman et al., 2012). More updated and direct
measurement based analysis of food waste in emerging and devel-
oping countries are badly needed.
In particular, the rising consumer food waste in China, the
world’s most populous and largest developing economy, deserves
a closer look due to its significant impact on the global food mar-
ket, resources use, and GHG emissions. China’s out-of-home food
consumption shows a dramatic growth in the past decade (State
Administration for Industry & Commerce of the People’s Republic
of China, 2009; Bai et al., 2010), as a result of its unique food cul-
ture and increasing household income. Consequently, consumer
food waste increases rapidly and repeatedly pops up in public
debates on food security, human health, and waste management
in China. For example, the illegal use of gutter oil from restaurant
food waste has long been headlines of many news stories; the pub-
lic’s concern on the enormous amount of food waste found in gov-
ernment paid banquets has led to a national wide campaign
against food waste that was initiated by the top leadership (Liu,
2014). Despite such growing public attention and media coverage,
there is still little information on the scales and patterns of restau-
rant food waste in China (Liu, 2014; Liu et al., 2013). Two rough
numbers were often blindly cited: on average about 10% of food
ordered ended up as waste in Chinese restaurants and restaurant
food waste in China per year would feed 200 million people
(Zheng, 2011). But a closer check shows that these data are purely
based on expert estimation with no empirical evidence.
We aim to address this data gap by making a first approxima-
tion of the current scales and patterns of restaurant food waste
in Chinese cities, based on field survey and direct weighing of food
waste from 3557 tables in 195 restaurants in four case cities in
China. Such an empirically determined and quantitative under-
standing will help inform the public, restaurant managers, and pol-
icy makers about the escalating scale of food waste and
consequently help explore strategies for food waste reduction.
The main objectives of this article are two folds: (i) to characterize
the total and per capita food waste generated in Chinese resultants
and their composition based on field survey and direct weighing in
the four case cities; and (ii) to explore the patterns of and reasons
behind restaurant food waste generation in China and their impli-
cations on food waste reduction strategies.
2. Materials and methods
2.1. System definition and sample selection
Consumer food waste can be found in different segments of the
consumer stage from restaurants to canteens to all other types of
Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
hospitality sectors. We focus only on the restaurant in this analysis,
due to the growing public attention it received and its rising con-
tribution to the generation of food waste in China (Liu, 2014). Food
waste is defined as the edible part that is left in the plate after the
meal, and the non-edible part such as food additives, flavorings,
cooking oil, and bones is not counted in our measurement.
Considering the varying cuisines and people’s consumer habits
in different parts of China, we selected four typical cities (Beijing,
Shanghai, Chengdu, and Lhasa) for case studies. Beijing is the cap-
ital of China and the largest city in northern China. Shanghai is the
economic center of China and one of the largest in the east.
Chengdu locates in western China and is the capital of Sichuan pro-
vince, hometown of one of the most popular Chinese cuisines.
Lhasa, the capital city of Tibet Autonomous Region of China, has
a unique food and culture tradition with an average altitude of
3650 m. More information and characteristics of the four case
cities were summarized in Table 1.
A stratified sampling method was used for the sample selection
of restaurants. We first analyzed data from the municipal Food and
Drug Administration, the main bureau of restaurant management
in China, about the number of all restaurants and the proportion
of different categories in different districts of each city. On the
basis of the data, we determined the sample districts in each city
and the number of restaurants of each category in the survey.

In China, restaurants are classified as different categories based
on their floor areas of business, i.e., large-size (more than
500m2), medium-size (between 150 and 500 m2), and small-
size (less than 150m2) restaurants. In addition, snack bars
which mainly provide refreshments, snacks, and fast food were
also selected in our survey. The number of surveyed restaurants
of each type (including snack bars) in different cities was sum-
marized in Fig. 1.

Three administrative districts within Beijing (Chaoyang District,
Haidian District, and Changping District), Shanghai (Changning
District, Minhang District, and Fengxian District), and Chengdu
(Wuhou District, Chengdu Hi-Tech Industrial Development
Zone, and Shuangliu County), were chosen based on a
downtown-to-suburban transect theory for sampling. Lhasa is
relatively small and thus we considered only the municipal
administrative district (Chengguan District) in the sampling.
Fig. 2 shows an example of Beijing for the total number of
restaurants by category and by district, which confirms the rep-
resentativeness of our sampled districts (e.g., Chaoyang, Haid-
ian, and Changping are the top three districts in terms of the
number of restaurants) and the share of different categories of
surveyed restaurants (when compared with Fig. 1).
In total, we have selected 195 restaurants in the four cities,
including 63 in Beijing, 51 in Shanghai, 54 in Chengdu, and 27 in
Lhasa. Most of Chinese people share all the food ordered on a table
when eating out, therefore we take each dining table as the basic
unit of our survey. For each restaurant, 20 tables of consumers
were randomly selected to conduct our survey. This all together
adds up to 3557 tables that were investigated in detail (Fig. 1). In
China, apart from tourism hotels which provide accommodation
together with a simple buffet breakfast, few restaurants provide
breakfast. Therefore, only lunch and dinner were included in our
survey.
2.2. Questionnaire design
For each table, two questionnaires were used to investigate the
amount and composition of food waste and consumer food waste
behaviors.
 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx 3

Table 1
Basic information on the four case cities (year 2015).

City Population of local Number of hosted GDP (billion Catering sales (billion Chinese Yuan) and % Feature of cuisines
residents (million) tourists (million) Chinese Yuan) in total retail sales of consumer goods
Beijing 21.70 270 2297 84.68 (8.2%) A variety of Chinese and foreign cuisines
Shanghai 24.15 276 2497 100.80 (10%; including hotels) A variety of Chinese and foreign cuisines
Chengdu 14.65 189 1080 59.46 (12%) A variety of cuisines, but dominated by local
Sichuan style which is popular across China
Lhasa 0.53 11.7 37.6 2.45 (11.9%; including hotels) A variety of cuisines, with a high share of
Tibetan style

Note: Data on population, GDP, catering and hotel sales, and total retail sales of consumer goods are from annual Statistical Bulletin of each city.

Fig. 1. The total number and share of surveyed restaurants, tables, and consumers in the four case cities.

8000
Snack bars and others Large-
awareness on environment protection, frugal habit, personal
Snack
7000 Small-size bars and size health, and food shortage (the first three perceptions assessed
others 12%
6000
Medium-size
24%
according to a 5-point Likert-type scale including ‘‘very weak,”
Large-size ‘‘weak”, ‘‘not sure”, ‘‘strong”, and ‘‘very strong”, while awareness
In total Medium-
5000 size of food shortage was tested by asking whether they believe there
35,693
Small- 32%
4000 size is still food shortage in China), and potential countermeasures to
32% reduce food waste. When there was more than one consumer on
3000
a table, this part was filled by the consumer who orders the meal
2000 or pays the bill for that table.
The second questionnaire was designed to record the amount
1000
and composition of food waste and it was completed by our trained
0 investigators. It consists of two parts. The first part is a weighing
table which includes the quantity of food offered and wasted in
the plate. The food was divided into 17 basic raw categories: rice,
wheat, maize, other grains, pork, beef, lamb, poultry, other meats,
Fig. 2. The number of restaurants by category and by district in Beijing. Data are eggs, aquatic products, vegetables, fruits, dairy products, beans,
obtained from the Food and Drug Administration of Beijing for the year 2013.
mushrooms and fungi, and potatoes and tubers. The second part
of this questionnaire is about the number of consumers on each
table and other qualitative information such as dining
The first questionnaire targets consumers and includes back-
environment.
ground information such as socio-demographic factors, consumers’

Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
4 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx
2.3. Field surveys and direct weighing
Our field surveys were conducted in July and August 2015, from
Monday to Sunday in order to capture the potential daily variances
of food consumption and waste. A direct weighing method was
used in the measurement. Electronic loading balances of 2 g to
5 kg were used to weigh the food and its containers. The specific
steps of direct weighing were as follows (see Fig. 3 for a few photos
in our survey):

Twenty sample tables were randomly selected in each restaurant.

The types of food raw materials and their proportions of each
dish were determined in the kitchen with the help of the chefs.

Each dish was weighed before service to determine the total
weight of food served (the food containers, e.g., bowls and
plates, were weighed beforehand and deducted here).

After the consumers finished their meal, questionnaire one was
filled by the person who ordered the meal or paid the bill.

All unfinished dishes left on the table were then collected after
the consumers left the table and then the food waste was sep-
arated and weighed.

These data were recorded in questionnaire two, together with
other basic information such as number of consumers on each
table observed by trained investigators.
2.4. Data treatment and analysis
The food waste per capita per meal (FW pcpm ) was then calcu-
lated by the following formula:
X
n X
d X
n
FW pcpm ¼ FW ij = Nn
i¼1 j¼1 i¼1
where FWij represents the waste amount of food category j on
table i, n is the number of tables surveyed in each city, d is the
number of food categories, and N is the number of surveyed con-
sumers on each table.
(a) Interviewing the chef
(c) Interviewing the consumer
Fig. 3. A few photos of the different steps of our survey (all photos taken by the authors).
Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
Based on the per capita food waste calculated above, we made a
further approximation of the annual restaurant food waste
(FWtotal) generated in each city as follows:
FW total ¼ FW pcpm pcpm
local  P local  F local þ FW tourist  P tourist  d  2:5 ð2Þ
where FW pcpm pcpm
local and FW tourist are the food waste per capita per
meal of local residents (defined as the population who lives in
the case city longer than six months) and tourists (from other parts
of China to the city), respectively (both derived from our survey).
Plocal represents the population of local residents and Ptourist repre-
sents the number of hosted tourists (person-times, obtained from
municipal Bureau of Trade and Industry). F local is the number of
times eating out in the restaurants for local residents (derived from
our questionnaire survey). And d is the average number of days
that tourists stay in a city (obtained from municipal Bureau of
Tourism) and we set the number of meals per tourism day as 2.5
(breakfast counted as half of lunch and dinner).
Microsoft Excel and SPSS 18.0 were adopted to analyze the data.
Descriptive analysis on total and per capita food waste in different
cities was firstly done. The differences in the amount of restaurant
food waste between tourists and local residents, among restaurant
categories, and with different purposes of meals were investigated
with independent sample t test and One-way ANOVA. The com-
monly used Tukey’s honest significant difference (HSD) test was
used for post hoc test. Linear regression analysis was employed
to examine the impact of consumers’ awareness on environment,
frugality, health, and food shortage on the amount of food waste.
Normal distribution was examined for skewness and kurtosis,
which was usually considered significant if the skewness coeffi-
ð1Þ
cient is smaller than 3 and kurtosis coefficient is smaller than 10
(Kline, 1998). If a normal distribution was rejected, data on food
waste per capita per meal were consequently normalized to meet
the ‘‘normality assumption” for the statistical techniques men-
tioned above. A p-value of 0.05 was used as the threshold of statis-
tical significance.
(b) Separating the food leftover
(d) Questionnaire filling and on-site weighing
 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx
3. Results
3.1. The amount of restaurant food waste in the four Chinese cities
The average amount of food waste per capita per meal in the
four cities is calculated as 93 g in 2015 (Fig. 4), among which veg-
etables contribute the most (29%), followed by rice (14%), aquatic
products (11%), wheat (10%), and pork (8%). The waste of all kinds
of meat (including pork, beef, mutton, poultry, and others) adds up
to 17% of the total food waste, while the waste of staple food
(including rice, wheat, maize, and other grains) accounts for 25%.
Fruits and dairy products are among the least wasted food cate-
gories, accounting for 1.1% and 0.2%, respectively.
When the four cities are compared (Fig. 5), the average amount
of restaurant food waste per capita per meal in Chengdu (103 g) is
the highest, while that of Beijing is the lowest (77 g), and Shanghai
(97 g) and Lhasa (98 g) are in between. Fig. 5 also shows the com-
position of food waste in the four case cities. Although vegetables
account for the biggest part in all four cities, its specific proportion
varies. In Lhasa the waste of vegetables accounts for 34% (or
33.3 g/cap/meal) of the total food waste, while in Shanghai the pro-
portion is only 25% (or 24.3 g/cap/meal). On the contrary, aquatic
products waste in Shanghai accounts for 16% (or 15.6 g/cap/meal)
of the total food waste, while that in Lhasa only accounts for
5.8% (or 5.7 g/cap/meal). The waste of pork per capita per meal
in Lhasa is much higher than that of the other three cities, account-
ing for 13% (or 13 g/cap/meal) of the total.
If these per capita per meal values were scaled up using Eq. (2),
the year 2015 total food waste in Shanghai, Beijing, Chengdu, and
Lhasa was 58.6  104 t, 41.7  104 t, 30.6  104 t, and 1.8  104 t,
respectively, which adds to approximately 1.3 million tons.
3.2. Variations by consumer groups, restaurant categories, and
purposes of meals
Our results show that food waste per capita per meal varies
considerably by factors such as consumer groups, restaurant cate-
gories, and purposes of meals.
Mushrooms &
fungi
2%
Fruits Potatoes &
Other grains 1%
0% Beans
Maize 7%
1%
Wheat
10%
Rice
14%
Fig. 4. Average food waste and composition of the four case Chinese cities.
Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
5
(1) Tourists versus local residents. The skewness (3.694 > 3)
and kurtosis (30.141 > 10) indicated that per capita food
waste did not follow a normal distribution, and it was thus
normalized for further analysis. As shown in Table 2, inde-
pendent sample test indicated that there was a significant dif-
ference between tourists and local residents in terms of food
waste per capita. The food waste quantity of tourists
(mean = 0.15, standard deviation = 0.90) was significantly
higher than that of residents (mean = 0.02, standard
deviation = 0.88).
Tourists in all four cities wasted on average 103 g/cap/meal of
food, higher than that of local residents (88 g/cap/meal). However,
the proportions of wasted food of these two groups show little dif-
ference (Fig. 6).
(2) Among different restaurant categories. In general, larger
restaurants generate more food waste per capita per meal
(Fig. 7). Consumers in large-size restaurants wasted on aver-
age 132 g/cap/meal which was nearly 3.5 times higher than
the amount in snack bars (38 g/cap/meal). Consumers in the
medium-size and small-size restaurants wasted 77 g and
69 g per capita per meal, respectively. In terms of the com-
position of wasted food, the four types of restaurants
showed a similar pattern. However, the shares of wheat
and poultry waste in snack bars were much higher than that
of the other three categories.
(3) Purposes of meals. Our results from one-way ANOVA with
post hoc tests (Turkey HSD) (see Table 3) showed that
restaurant food waste from friends gathering (107 g) and
business banquet (102 g) were significant higher than that
of the other three groups (95 g, 90 g, 67 g for family gather-
ing, working meal, and personal dining, respectively) (F
(4, 3141) = 37.63, p < 0.05). Composition wise (Fig. 8), work-
ing meals led to more rice waste (20% of the total, or
18 g/cap/meal), personal dining led to a higher share of
wheat waste (14% of the total, or 9.6 g/cap/meal), and
business banquet resulted in a higher waste of aquatic
Dairy products
0%
Beef
Pork 3%
Lamb
8%
tubers 0%
5%
Poultry
6%
Aquatic
On average products
93 g/cap/meal 11%
Other
meats
0%
Eggs
2%
Vegetables
29%
6 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx

Pork
Beef
Lhasa 13,2 5,7 33,3 9,8 8,5 98 Lamb
Poultry
Aquatic products
Other meats
Chengdu 9,1 11,1 31,4 13,9 5,5 103
Eggs
Vegetables
Rice
Shanghai 7,9 15,6 24,3 14,8 9,1 Wheat
97
Maize
Other grains
Fruits
Beijing 4,1 7,7 21,1 11,3 13,0 77 Beans
[g/cap/meal] Mushrooms & fungi
Potatoes & tubers
0 10 20 30 40 50 60 70 80 90 100 110 Dairy products

Fig. 5. Food waste per capita per meal and proportions of each case city.

Table 2
Statistical comparison of food waste between tourists and residents.

Group
Tourists Local residents t df
Food waste per capita per meal (normalized) 0.15 (0.90) 0.02 (0.88) 2.64** 3021
* **
Note: p < 0.05, p < 0.01. The values appearing in parentheses are standard deviations. df means degree of freedom.

35
[g/cap/meal]
30
Tourist [103 g/cap/meal]
25
Resident [88 g/cap/meal]
20

15

10

0
Wheat

Beans

Potatoes &
Pork

Beef

Lamb

Poultry

products
Other meats

Other grains

Fruits

Mushrooms

products
Eggs

Vegetables

Rice

Maize
Aquatic

Dairy
& fungi

tubers

Fig. 6. The comparison of restaurant food waste per capita per meal between tourists and local residents.

products (19% of the total, or 19 g/cap/meal) and beef (5.3% they do ‘‘occasionally”, and 5% said ‘‘never”. When asked the rea-
of the total, or 5.4 g/cap/meal). sons why they did not pack the food left on the restaurant table,
(4) Dinners versus lunches. The amount of average food waste ‘‘cheap price”, ‘‘health considerations”, and ‘‘inconvenience for
per capita per meal for dinner is 104 g, which is higher than packing” were the most often mentioned. Even though the food
that of lunch (89 g). left-over was packed and took back home, it may also be eventu-
ally wasted. Good news is that more than 82% of respondents said
3.3. Results on consumers’ awareness on environment, frugality, that they do eat the taken back food left-over, while 10% confessed
health, and food shortage and thoughts on food waste reduction they feed pets with the doggie food bags, and 8% eventually throw
it away into bins.
Table 4 shows that the linear regression model was significant, For the multi-choices questions on the measures to reduce
revealing a significant linear relationship between the independent restaurant food waste, 56% respondents argued the restaurants
and dependent variables (F(3, 3288) = 4.826, p < 0.05). Awareness should ‘‘serve the food with smaller quantity” and 51% agreed that
of frugal habit relates significantly to the amount of food waste, a transparent description and reminder of the food quantity during
whereas the awareness on environmental protection, personal ordering would help reduce waste. In addition, ‘‘improving the
health, and food shortage appears non-significant. quality of food” and ‘‘punishing the behavior of food waste” were
Regarding the habit of packaging food left-over after meals, 43% other most recommended measures by the respondents.
of respondents indicated that they ‘‘often” do so, 53% indicated

Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx 7

140
[g/cap/meal]
Dairy products Potatoes & tubers Mushrooms & fungi
Beans Fruits Other grains
120 10,9 Maize Wheat Rice
Vegetables Eggs Other meats
6,9
100 Aquatic products Poultry Lamb
15,4 Beef Pork

80

5,2
38,5
5,0
60 9,1
11,8
12,3

40 12,1
19,2 1,5
23,6
20,2 12,2
20 6,6
7,1 1,7 6,3
11,7 0,9
5,7 7,4
0 1,7
Large-size Medium-size Small-size Snack bar

Fig. 7. The comparison of restaurant food waste per capita per meal among different categories of restaurants.

Table 3
One-way ANOVA for mean differences in food waste per capita per meal for different purposes of meals.

N Mean (SD) df F-value

Food waste per capita per meal (normalized) Business banquet 120 0.22 (0.88) Between groups 4 37.63**
Family gathering 649 0.11 (0.90) Within groups 3141
Friends gathering 918 0.29 (0.89) Total 3145
Working meal 540 0.08 (0.87)
Personal dining 919 0.17 (0.84)
Total 3146

Note:
**
p < 0.01.

35 [g/cap/meal]

30

25
Dairy products
Potatoes & tubers
20 Mushrooms & fungi
Beans
Fruits
Other grains
15 Maize
Wheat
Rice
10 Vegetables
Eggs
Other meats
5 Aquatic products
Poultry
Lamb
0 Beef
Pork

Fig. 8. The comparison of restaurant food waste per capita per meal with different purposes.

Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
8 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx

Table 4
Regression results between food waste and consumers’ awareness on environment protection, personal health, frugal habit, and food shortage.

b (unstandardized coefficient) b (standardized coefficient) Statistic t-value Statistic p-value

Dependent variable (food waste per capita per meal) Independent variable (awareness on)
Environmental protection 0.000 0.000 0.012 .990
Personal health 0.008 0.009 0.394 0.694
Frugal habit 0.053 0.063 3.103 0.002**
Food shortage (1 = Yes, 0 = No) 0.061 0.032 1.802 0.072

Model statistics (R2 = 0.004, Adj. R2 = 0.003, F(4, 3094) = 3.305, p < 0.05).
*
p < 0.05.
**
p < 0.01.

Fig. 9. Comparison of restaurant food waste in Chinese cities with that in developed countries. GDP data are from World Bank. Restaurant food waste data source: Japan,
2000, 2007, 2008, 2009, 2010, 2011, 2012 (Parry et al., 2015); China, 2008 (Gao et al., 2013); Sweden, 2010 (Buchner et al., 2012; Eriksson, 2012; Blomgren and Bylund, 2013),
Sweden, 2011 (Gjerris and Gaiani, 2013), Sweden, 2012 (Marthinsen et al., 2012; Eriksson, 2015); Germany, 2012 (Blanke, 2014), Germany, 2013 (Blanke, 2015); Finland,
2010, 2012 (Silvennoinen et al., 2012), Finland, 2013 (Eriksson, 2015); Norway, 2012 (Marthinsen et al., 2012); and USA, 2003 (Jones, 2005).

4. Discussion
4.1. Implications and uncertainties of our results
A clear understanding of the quantity, patterns, and reasons of
food waste generation is a prerequisite to explore and prioritize
strategies for food waste reduction. Such an understanding
remains extremely poor or outdated for the consumer stage in
developing countries, which are also experiencing a fast urbaniza-
tion process and dietary transition. Our surveys provide a first
approximation of the scale of restaurant food waste in Chinese
cities, a sector that deserves special attention due to the special
and long dietary culture and increasing household income in
China. This information is paramount to setting targeted interven-
tions to prevent and reduce avoidable food waste in restaurants as
well as providing a baseline for national food waste reduction tar-
get. Such a large-scale and first-hand data collection based effort
on consumer food waste is, to our best knowledge, among the first
of its kind in China. It can be duplicated in other cities in China and
thus contribute to a better understanding of the scales, patterns,
and impacts of consumer food waste in China and other emerging
and developing countries.
On average, 93 g of food per capita per meal was wasted in the
four case cities in 2015. This equals to approximately 11 kg/cap/
year, if the average values of local residents in the four case cities
were scaled up to an annual basis. When compared to existing data
of restaurant food waste in western countries (Fig. 9), China’s cur-
rent level of waste is already close to the average (between 5 and
15 kg/cap) of Nordic countries, although per capita GDP of China is
still much lower than that of western countries. The four case cities
generated in total approximately 1.3 million tons of restaurant
food waste in 2015, which is already higher than the total hospital-
ity food waste of UK in 2009 (Environmental Agency North East,
2009) and is close to the total restaurant food waste of Germany
(1.9 million tons) in 2013 (Blanke, 2015).
China has 22% of the world’s population but only 7% of the
world’s arable land. Therefore, food security has been and will
remain as a priority for China. While increasing primary food pro-
duction is paramount to meet the future increase in final demand,
tensions between production and access to food can also be
reduced by tapping into the potential reduction of food waste
(Liu et al., 2016). In the foreseen future, China’s continuing increase
of economy and household income, growth of urban population,
and development of tourism will result in a continuing booming
catering and hospitality industry. This trend also means, if no effec-
tive actions were taken, more food would be wasted in restaurants.
Considering China’s important role in achieving global sustainabil-
ity, such a growing trend of food waste will raise enormous con-
cerns on food security and environmental sustainability not only
for China, but also for the whole world.
As a first exploratory study, our results bear some uncertainties
and limitations that should be mentioned and addressed in the
future.

First, due to time and cost constraints, we covered only four
case cities and three districts within each city in our survey.
The sampling of restaurant was based on a rough understanding
of restaurant categories and the surveyed tables in each restau-

Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx
rant were randomly selected. Therefore our results may not be
representative for all China and do not sufficiently reflect poten-
tial differences among cities and between urban and rural areas.

Second, our data measured by direct weighing and field survey
could not avoid intrinsic inconsistency and inaccuracy. For
example, food composition in each dish was estimated with
the help of chefs and may not be accurate. The varying experi-
ence and stringency of student investigators may also increase
the data uncertainty (e.g., inconsistent understanding of non-
edible parts).

Third, the primary goal of our survey is for a first estimation of
the scale of restaurant food waste in Chinese cities. Some pat-
terns (e.g., by restaurant category and purpose of meals) were
discussed, but the collected information was not detailed
enough for any further in-depth analysis of many other impor-
tant factors, for example, the role of portion size, seasonal or
daily variances, and differences between Chinese food and
western food.

Fourth, we used tables as the basic unit in our survey and calcu-
lation, since Chinese mostly share food and it is difficult to mea-
sure food waste individually. So the average values reflect each
table rather than behavior of each person. Questionnaire one
was filled only by the consumer who orders the food or pays
the bill but not all the consumers on the table. This may con-
tribute to the low correlation in our regression analysis
between ordering consumer’s awareness and per capita food
waste generation of the entire table.

Last, our calculation did not include food waste in the restau-
rant kitchens during food processing and cooking, so the results
are an underestimation.
4.2. Patterns of and mitigation strategies for restaurant food waste in
Chinese cities
When results of the four cities are compared, it is not surprising
that consumers in Chengdu and Lhasa wasted more food per capita
per meal than consumers in Beijing and Shanghai. Chengdu has been
considered as the most relaxing city and popular recreation destina-
tion in China, with a dominating hot pot cuisine (suitable for people
eating together for long time) and dining-out culture. This makes
Chengdu the most wasteful among the four case cities regarding
restaurant food waste. On the contrary, as the capital of China, Bei-
jing could have been affected by the recent political crack-down
on food waste (Liu, 2014) and campaigns like ‘‘Clean Your Plate”
laughed by the government may have made an impact as well.
The difference of restaurant food waste between local residents
and tourists in our study confirms early claim on the uniqueness of
tourism catering behavior (Mak et al., 2012a, 2012b). Tourists’ food
consumption occurs usually in a foreign and unfamiliar context.
Most tourists have no ideas about the quantity and quality of the
food when they order, which may consequently result in more
waste. For example, the high level of restaurant food waste in
Lhasa can be partly explained by its role as a major tourism city
in western China. Many tourists lose their appetites because of
the high altitude (3650 m above sea level) and their unfamiliarity
with the local Tibetan food.
The purposes of meals have a significant impact on the amount
of restaurant food waste. For example, business banquet and
friends gathering, which also often take place in the large or med-
ium size restaurants, result in more food waste than the other
types. Specifically, more than 57% of business banquet and 41%
of friends gathering occurred in large size restaurants. Over half
of working meal and private dining, on the contrary, took place
in small size restaurants or snack bars. In addition, 76% of working
meals and 64% of private dining are in the form of lunches which
also wasted relatively less food than dinners.
Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007
9
Our results also demonstrate the role of culture in food waste
generation (Stuart, 2009; Liu, 2014). For example, consumers in
Shanghai wasted much more aquatic products than other cities
because Shanghai is a coastal city and the Shanghai cuisine has a
long history in the use of aquatic products; On the contrary, the
share of aquatic products waste in Lhasa is very low because Tibe-
tans do not eat them for cultural and religion reasons and hence it
is rarely provided by local restaurants. In general, over-ordering is
seen as a kind of hospitality in the Chinese culture (the culture of
‘‘mianzi”). If all the food on the table is finished after a meal, it
could be interpreted as the host does not prepare or order enough
food. So it’s not surprising to observe that business banquet wasted
more food, and from snack bars to large size restaurants, con-
sumers focus less and less on food but more and more on the social
relationship. When the food was considered as the social tool
rather than delicacy or necessity, food waste will be the last thing
people are concerned about.
Food waste is not only ethically unacceptable, but has also sig-
nificantly adverse impacts on human health, food security,
resource use, and the environment. The results of our pilot study
reveal that restaurant food waste in Chinese cities is already at a
high level and calls for particular attention and immediate actions.
Based on the food waste hierarchy (Papargyropoulou et al.,
2014), improved food waste disposal system is badly needed in
China. Food waste takes up a high share (ranging from 50% to
70%) in municipal solid waste in Chinese cities (Tai et al., 2011)
and most of food waste is mixed with other municipal solid waste
streams and eventually incinerated or landfilled (for example, on a
national average 56.6% was landfilled in 2009) (Hu et al., 2012).
Other strategies such as recovering energy via anaerobic digestion
of food waste, recycling food waste via compositing and/or into
animal feed, and redistributing food surplus via food banks should
also be explored.
Food waste prevention is a fundamental but also difficult strat-
egy, and it requires involvement of multiple stakeholders. Con-
sumers play a crucial role in plate waste reduction (for example,
our results show that consumers with stronger frugality awareness
tend to waste less food), and their actual or anticipated food per-
ceptions and food purchase behaviors also influence other stake-
holders (e.g., the restaurants). For the catering business, the
pursuit of profit will encourage them to improve the efficiency of
processing and utilization of food in the kitchen. However, they
lack economic driver to reduce consumer food waste, because
the more consumers order, the more money they will make.
Governmental supervision and legislative guidance play a particu-
larly important role, especially for China. For example, Chinese
government’s recent crack-down on official extravagance and gov-
ernmental reception meals at public expenses has shown an imme-
diate impact on restaurant food waste reduction. More bottom-up
initiatives from non-governmental organizations and social cam-
paigns such as the ‘‘Clean Your Plate” initiative should also be
encouraged.
Acknowledgements
This work is funded by the Natural Science Foundation of China
(71233007; 41301141) and National Key Research and Develop-
ment Plan of China (2016YFE0113100). Ling-en Wang thanks China
Postdoctoral Science Foundation (2014M561041; 2015T80128)
and China Scholarship Council. Gang Liu acknowledges the Inter-
national Network Programme funding of the Danish Agency for
Science, Technology and Innovation (5132-00029B and 6144-
00036). We thank Li Xue, Xiaochang Cao, Liwei Gao, Dan Zhang,
Xuewang Dong, and all other team members and student volun-
teers for data collection and Junfei Bai, Wen Yu, Shiwei Xu, Yifei
Zhang, and Jun Liu for their advice in our survey.
10 L.-e. Wang et al. / Waste Management xxx (2017) xxx–xxx

References Kummu, M., de Moel, H., Porkka, M., Siebert, S., Varis, O., Ward, P.J., 2012. Lost food,
wasted resources: global food supply chain losses and their impacts on
freshwater, cropland, and fertilizer use. Sci. Total Environ. 438, 477–489.
Aschemann-Witzel, J., de Hooge, I., Amani, P., Bech-Larsen, T., Oostindjer, M., 2015.
Jones, T.W., 2005. Using Contemporary Archaeology and Applied Anthropology to
Consumer-related food waste: causes and potential for action. Sustainability 7
Understand Food Loss in the American Food System. Bureau of Applied
(6), 6457–6477.
Research in Anthropology, University of Arizona, Tucson, Arizona.
Bai, J., Wahl, T.I., Lohmar, B.T., Huang, J., 2010. Food away from home in Beijing:
Liu, G., 2014. ‘‘Food Losses and Food Waste in China: A First Estimate”, OECD Food,
effects of wealth, time and ‘‘free” meals. China Econ. Rev. 21, 432–441.
Agriculture and Fisheries Papers, No. 66, OECD Publishing. doi: http://dx.doi.
Banks, G., Collison, R., 1981. Food waste in catering. Proc. Inst. Food Sci. Technol. 14,
org/10.1787/5jz5sq5173lq-en.
181–189.
Liu, G., Liu, X., Cheng, S., 2013. Food security: curb China’s rising food wastage.
Blanke, M.M., 2014. Reducing ethylene levels along the food supply chain – a key to
Nature 498, 170.
reducing food waste? J. Sci. Food Agric. 94 (12), 2357–2361.
Liu, Y., Cheng, S., Liu, X., Cao, X., Xue, L., Liu, G., 2016. Plate waste in school lunch
Blanke, M., 2015. Challenges of reducing fresh produce waste in Europe: from farm
programs in Beijing, China. Sustainability 8, 1288.
to fork. Agriculture 5 (3), 389–399.
Lo, J., Jacobson, A., 2011. Human rights from field to fork: Improving labour
Blomgren, M., Bylund, J., 2013. The Connection Between the Issue of Food Waste
conditions for food-sector workers by organizing across boundaries. Race/
and its Collection for Biogas: A Case Study of the Municipality of Stockholm
Ethnicity Multidiscipl. Global Contexts 5, 61–82.
Master Thesis. Stockholm University, Stockholm.
Mak, A.H., Lumbers, M., Eves, A., Chang, R.C., 2012a. Factors influencing tourist food
Buchner, B., Fischler, C., Gustafson, E., Reilly, J., Riccardi, G., Ricordi, C., Veronesi, U.,
consumption. Int. J. Hospital. Manage. 31 (3), 928–936.
2012. Food Waste: Causes, Impacts and Proposals. Barilla Center for Food &
Mak, A.H., Lumbers, M., Eves, A., 2012b. Globalisation and food consumption in
Nutrition, Parma, Italy.
tourism. Annals Tourism Res. 39 (1), 171–196.
Charlebois, S., Creedy, A., von Massow, M., 2015. ‘‘Back of house” – focused study on
Marthinsen, J., Sundt, P., Kaysen, O., Kirkevaag, K., 2012. Prevention of Food Waste in
food waste in fine dining: the case of Delish restaurants. Int. J. Cult. Tourism
Restaurants, Hotels, Canteens and Catering. Nordic Council of Ministers,
Hospital. Res. 9 (3), 278–291.
Copenhagen, Denmark.
Dorward, L.J., 2012. Where are the best opportunities for reducing greenhouse gas
Monier, V., Shailendra, M., Escalon, V., O’Connor, C., Gibon, T., Anderson, G.,
emissions in the food system (including the food chain)? A comment. Food
Hortense, M., Reisinger, H., 2011. Preparatory study on food waste across EU 27.
Policy. 37, 463–466.
European Commission (DG ENV) Directorate C-Industry. 2010. ISBN: 978-92-
Eriksson, M., 2012. Retail Food Wastage: A Case Study Approach to Quantities and
79-22138-5.
Causes Master Thesis. Swedish University of Agricultural Sciences, Uppsala,
Nahman, A., de Lange, W., Oelofse, S., Godfrey, L., 2012. The costs of household food
Sweden.
waste in South Africa. Waste Manage. 32 (11), 2147–2153.
Eriksson, M., 2015. Supermarket Food Waste: Prevention and Management with the
Neff, R.A., Spiker, M.L., Truant, P.L., 2015. Wasted food: US consumers’ reported
Focus on Reduced Waste for Reduced Carbon Footprint Ph.D. Dissertation.
awareness, attitudes, and behaviors. PLoS One 10 (6), e0127881.
Uppsala University, Uppsala.
Papargyropoulou, E., Lozano, R., Steinberger, J.K., Wright, N., Ujang, Z.B., 2014. The
Eshel, G., Shepon, A., Makov, T., Milo, R., 2014. Land, irrigation water, greenhouse
food waste hierarchy as a framework for the management of food surplus and
gas, and reactive nitrogen burdens of meat, eggs, and dairy production in the
food waste. J. Clean. Prod. 76, 106–115.
United States. Proc. Natl. Acad. Sci. USA 111, 11996–12001.
Parfitt, J., Barthel, M., Macnaughton, S., 2010. Food waste within food supply chains:
Environmental Agency North East, 2009. A Study of Public Sector Food Waste
quantification and potential for change to 2050. Philos. Trans. Roy. Soc. B Biol.
Arisings and Processing Options within the North East Region.
Sci. 365 (1554), 3065–3081.
FAO, 2012. The State of Food Insecurity in the World 2012. Economic Growth is
Parfitt, J., 2013. Global food waste campaigns suffer from data deficiency [WWW
Necessary But Not Sufficient to Accelerate Reduction of Hunger and
Document]. <http://www.theguardian.com/global-development-professionals-
Malnutrition. Food and Agriculture Organization of theUnited Nations (FAO),
network/2013/oct/28/global-food-waste-tesco-fao-data> (accessed 22.02.17).
Rome, Italy.
Parry, A., Bleazard, P., Okawa, K., 2015. Preventing food waste: case studies of Japan
Gao, L., Cheng, S., Cao, X., Zhang, D., Liu, X., Qin, Q., Liu, Y., 2013. An overview of the
and the United Kingdom. OECD Food Agriculture & Fisheries Papers, No. 76,
resources and environmental issues from wasted food in urban catering across
OECD Publishing, Paris.
China. J. Resour. Ecol. 4 (4), 337–343.
Silvennoinen, K., Katajajuuri, J.M., Hartikainen, H., Jalkanen, L., Koivupuro, H.K.,
Gjerris, M., Gaiani, S., 2013. Household food waste in Nordic countries: estimations
Reinikainen, A., 2012. Food waste volume and composition in the Finnish
and ethical implications. Nord. J. Appl. Ethics. 7 (1), 6–23.
supply chain: special focus on food service sector. In: Fourth International
Godfray, H.C.J., Beddington, J.R., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F.,
Symposium on Energy from Biomass and Waste. CISA Publisher, Venice, Italy.
Toulmin, C., 2010. Food security: the challenge of feeding 9 billion people.
State Administration for Industry & Commerce (SAIC) of the People’s Republic of
Science 327 (5967), 812–818.
China. 2009. National catering Industry Development Plan (2009–2013).
Gustavsson, J., Cederberg, C., Sonesson, U., Van Otterdijk, R., Meybeck, A., 2011.
Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M., De Haan, C., 2006.
Global Food Losses and Food Waste. Food and Agriculture Organization of the
Livestock’s Long Shadow: Environmental Issues and Options. FAO Publishing,
United Nations, Rome, Italy.
Rome.
Grey, C.E., Dubois, R.A.A., 1947. A study of plate waste in the college dining room. J.
Stuart, T., 2009. Waste: Uncovering the Global Waste Scandal. Penguin, London, UK.
Am. Diet. Assoc. 23, 330–332.
Tai, J., Zhang, W., Che, Y., Feng, D., 2011. Municipal solid waste source-separated
Gruber, L.M., Brandstetter, C.P., Bos, U., Lindner, J.P., Albrecht, S., 2015. LCA study of
collection in China: a comparative analysis. Waste Manage. 31 (8), 1673–1682.
unconsumed food and the influence of consumer behavior. Int. J. Life Cycle
Thi, N.B.D., Kumar, G., Lin, C.Y., 2015. An overview of food waste management in
Assess., 1–12
developing countries: current status and future perspective. J. Environ. Manage.
Hu, X., Zhang, M., Yu, J., Zhang, G., 2012. Food waste management in China: status,
157, 220–229.
problems and solutions (in Chinese). Acta Ecol. Sin. 32 (14), 4575–4584.
Youngs, A.J., Nobis, G., Town, P., 1982. A Report to the Minister of Agriculture,
Kline, R.B., 1998. Principles and Practice of Structural Equation Modeling. Guilford
Fisheries and Food on an Investigation into Food Waste in the Hotel Sector of
Press, New York.
the Catering Industry. Dorset Institute of Higher Education, Poole.
Koivupuro, H.K., Jalkanen, L., Katajajuuri, J.M., Reinikainen, A., Silvennoinen, K.,
Youngs, A.J., Nobis, G., Town, P., 1983. Food waste from hotels and restaurants in the
2010. MTT report food waste in the supply chain. Elintarvikeketjussa syntyvä
UK. Waste Manage. Res. 1 (4), 295–308.
ruokahävikki, kirjallisuuskatsaus 12, 73 http://www.mtt.fi/mttraportti/pdf/
Zheng, C., 2011. Serious food waste found in the catering sector in China (China
mttraportti12.pdf.
Daily News, in Chinese) [WWW Document]. <www.chinadaily.com.cn/hqgj/
jryw/2011-03-08/content_1957714.html> (accessed 22.02.17).

Please cite this article in press as: Wang, L.-e., et al. The weight of unfinished plate: A survey based characterization of restaurant food waste in Chinese
cities. Waste Management (2017), http://dx.doi.org/10.1016/j.wasman.2017.04.007