Waste Management 60 (2017) 32–41

Contents lists available at ScienceDirect

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

Does recyclable separation reduce the cost of municipal waste

management in Japan?
Rosaria Chifari a,1, Samuele Lo Piano a,1, Shigeru Matsumoto b,1,⇑, Tomohiro Tasaki c,1
a
Institute for Environmental Science and Technology, Universitat Autònoma de Barcelona, Edifici Z, 08193 Bellaterra, Cerdanyola del Vallès, Barcelona, Spain
b
Department of Economics, Aoyama Gakuin University, Room 828 Building 8, 4-4-25 Shibuya, Shibuya, Tokyo 150-8366, Japan
c
Center for Material Cycles and Waste Management Research, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan

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

Article history: Municipal solid waste (MSW) management is a system involving multiple sub-systems that typically
Received 17 October 2016 require demanding inputs, materials and resources to properly process generated waste throughput.
Revised 11 January 2017 For this reason, MSW management is generally one of the most expensive services provided by munici-
Accepted 11 January 2017
palities. In this paper, we analyze the Japanese MSW management system and estimate the cost elasticity
Available online 22 January 2017
with respect to the waste volumes at three treatment stages: collection, processing, and disposal.
Although we observe economies of scale at all three stages, the collection cost is less elastic than the dis-
Keywords:
posal cost. We also examine whether source separation at home affects the cost of MSW management.
Municipal solid waste management cost
Japan
The empirical results show that the separate collection of the recyclable fraction leads to reduced pro-
Source separation cessing costs at intermediate treatment facilities, but does not change the overall waste management
Treatment stage cost. Our analysis also reveals that the cost of waste management systems decreases when the service
is provided by private companies through a public tender. The cost decreases even more when the service
is performed under the coordination of adjacent municipalities.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction et al., 2012), municipalities in developing countries have started

Issues of solid waste management are expected to become a
major environmental problem in the near future for cities world-
wide (Saeed et al., 2009). For instance, the World Bank (2012)
expects the volume of waste globally generated in urban contexts
to increase from the 2009 annual figure of 1.3 billion tons to
2.2 billion tons in 2025.
In recent years, the majority of municipalities in developed
countries have implemented various forms of programs and
schemes to promote waste reduction activities. Some municipali-
ties have developed recycling programs to collect recyclable mate-
rials from trash while also enhancing nutrient recovers from bio-
degradable waste (Zabaleta and Rodic, 2015). Others municipali-
ties have adopted pay-as-you-throw programs to educate residents
on the costs of waste disposal (Kinnaman, 2006; De Jaeger and
Eyckmans, 2015) and to promote community efforts towards
waste-throughput reduction (Castagna et al., 2013; Di Leo and
Salvia, 2017). Although it is known that traditional reuse practices
have helped save substantial amounts of reusable resources (Bari
⇑ Corresponding author.
E-mail address: shmatsumoto@aoyamagakuin.jp (S. Matsumoto).
1
The list of authors is alphabetical and reflects equal contribution.
implementing waste management programs analogous to those
used in developed countries (Troschinetz and Mihelcic, 2009;
Sarkhel et al., 2016) due to a shortage of waste-treatment
capacities.
Despite the development of the above-mentioned waste reduc-
tion and recycling programs, solid waste management schemes
can still be listed as some of the most expensive services provided
by municipalities. In low-income countries, public spending on this
service typically amounts to 20–50 percent of the recurrent avail-
able budget (World Bank, 2011). The allocation of such a relevant
fraction of public resources makes it difficult for these countries
to address citizens’ remaining basic needs. In such stressed finan-
cial conditions and in view of increasingly aging population-
driven social security expenditures, the reduction of MSW man-
agement costs is becoming a major policy challenge for developed
countries.
The MSW management cost structure has been frequently ana-
lyzed in the literature. However, we believe there are at least two
limitations to previous studies. First, the ways in which waste sep-
aration affects costs of MSW management have not been exam-
ined; second, cost elasticity with respect to waste volume at the
collection, processing, and disposal stages has not been adequately
covered. With the aim of addressing these two research gaps, we

http://dx.doi.org/10.1016/j.wasman.2017.01.015
0956-053X/Ó 2017 Elsevier Ltd. All rights reserved.
 R. Chifari et al. / Waste Management 60 (2017) 32–41
Nomenclature
Roman symbols
QT quantity of total waste reported in the survey [ton]
QR quantity of waste adjusted for costs of treatment [ton]
A constant [–]
L total quantity of labor inputs [–]
K total quantity of capital inputs [–]
w wage [–]
r rental cost of capital [–]
CT total cost of waste management [thousands of dollars]
CC cost at collection stage [thousands of dollars]
CP cost at processing stage [thousands of dollars]
CD cost at disposal stage [thousands of dollars]
c0 constant [thousands of dollars]
c1 cost elasticity with respect to wage [%]
analyze the Japanese MSW management system in this study. The
case of Japan is particularly suitable to study as all Japanese munic-
ipalities file their waste management information in a specified
and standardized format reported annually to the central govern-
ment. This standardized data collection system allowed us to carry
out a cost structure analysis in a systematic manner. To this extent,
we believe that our empirical results could prove useful not only
for MSW management planning in Japan but also for such planning
in other countries. By analyzing cost variations of each parameter,
it is possible to plan the most relevant course of action needed to
lower management costs. Such an approach could prove particu-
larly useful for policy makers when planning local management
systems.
The remainder of the paper is organized as follows. Section 2
reviews the empirical findings of previous studies. In Section 3,
we summarize the data used and described the waste management
system in Japan. In Section 4, we specify our empirical models. Sec-
tion 5 presents the empirical results of our statistical analysis. Sec-
tion 6 reports our conclusions and policy implications.
2. Literature review
The MSW management cost structure has been analyzed in the
literature for the last 50 years due to its significance and to its
specific importance to overall systems planning. Table 1 summa-
rizes the empirical findings of the nine most-frequently cited arti-
cles in scientific journals written on this subject.
The table presents geographic information and the size of each
municipal sample examined in each study. Each row represents
factors taken into account to determine their influence on the over-
all costs of municipal solid waste management schemes.
Hirsch (1965) initiated a study on MSW disposal by analyzing
data from 24 municipalities in the St. Louis county area in 1960.
Since then, other early studies such as Kitchen (1976), Stevens
(1978) and Dubin and Navarro (1988) have been conducted in
North America. In contrast, more recent studies have been per-
formed in European countries (Italy (Greco et al., 2015), the
Netherlands (Dijkgraaf and Gradus, 2003), Spain (Bel and Fageda,
2010), Sweden (Ohlsson, 2003), and the United Kingdom
(Szymanski and Wilkins, 1993; Szymanski, 1996)).
Several influential factors examined in this study have also been
individually considered in the literature. For instance, the
economy-of-scale aspect has been examined in Dubin and
Navarro (1988), Szymanski and Wilkins (1993), Szymanski
(1996), Bel and Fageda (2010) and Greco et al. (2015), and these
studies revealed increasing returns to scale (IRS) in MSW manage-
33
c2 cost elasticity with respect to the quantity of total waste
[%]
X vector of control variables [–]
B vector of the coefficient for control variables [–]
k waste category [–]
m total number of waste categories [–]
Greek symbols
a labor coefficient in the Cobb-Douglas production func-
tion [–]
b capital coefficient in the Cobb-Douglas production func-
tion [–]
sk share of the kth category of waste [0–1]
hk cost elasticity with respect to the waste share [%]
ment production schemes. Collins and Downes (1977), Callan and
Thomas (2001), Reeves and Barrow (2000) and Dijkgraaf and
Gradus (2003) found constant returns to scale (CRS) while
Domberger et al. (1986) and Bosch et al. (2000) found decreasing
returns to scale (DRS).
In addition, in regards to market-structure factors, Tickner and
McDavid (1986) found that the costs of MSW management
decrease when municipalities contract out their services to a pri-
vate waste management company. Moreover, the impact of com-
pulsory competitive tendering on the costs of MSW management
in the UK was estimated by Domberger et al. (1986). The authors
found no difference in public and private costs under competitive
contracting but that public costs become higher than private costs
in the absence of competition. Zhu et al. (2016) evaluated the per-
formance of a hypothetical public-private partnership for the city
of Hamilton (Ontario, Canada) in comparison with public and pri-
vate collection systems and reached the conclusion that the
mixed-management option could serve as a reasonable solution.
Estimations of recycling costs have also been attempted by
Kinnaman et al. (2014) who assumed that the social cost of recy-
cling can be expressed as the sum of municipal costs and revenues,
recycling costs to households, external disposal costs, and external
recycling benefits. The authors estimated an optimum recycling
rate, i.e., the rate that minimizes social costs, of roughly 10% for
Japan and concluded that the value of citizen engagement in recy-
cling programs is actually overestimated. Callan and Thomas
(2001) and Weng and Fujiwara (2011) estimated the role of waste
and recycling cost functions and came to the exact opposite con-
clusion that recycling leads to the reduction of total waste manage-
ment costs. Massarutto et al. (2011) performed an economic life-
cycle analysis to compare the costs of alternative waste manage-
ment scenarios in north-central Italy. According to the authors’
findings, the financial costs of waste management increase when
waste separation at the household level increases beyond 60%
and by only 30% for some municipalities. Greco et al. (2015) also
examined the impacts of source separation on the costs of a com-
plete MSW management system. They found that the overall per-
centage of recycled waste is positively related to paper,
paperboard, and organic collection costs.
While these papers have examined MSW management from
various angles, we believe that the literature still presents several
limitations first with respect to geographical breadth, as studies
conducted outside of North America and Europe remain very lim-
ited. One study by Ren and Hu (2014) analyzed the financial per-
formance of MSW services in China and found that labor
accounts for more than half to up to three-quarters of the opera-
 34
Table 1
A summary of the empirical findings on the cost structure of MSW management.

Hirsch Stevens Dubin and Szymanski and Wilkins (1993), Callan and Dijkgraaf and Ohlsson Bel and Fageda Greco et al.
(1965) (1978) Navarro (1988) Szymanski (1996) Thomas (2001) Gradus (2003) (2003) (2010) (2015)
Country USA USA USA UK USA Netherlands Sweden Spain Italy
Number of 24 340 261 365 110 85 115 65 67

R. Chifari et al. / Waste Management 60 (2017) 32–41

municipalities
Type of cost Average disposal cost Total Average Total cost Total disposal and Total Total cost Total cost Cost per
per pickup collection cost disposal cost recycling cost collection cost per ton inhabitant
Economies of scale Not observed Decreasea Decreasea Decrease Not observed Not observed Not Decreasea Decreaseb
observed
Economies of density Decrease Decreasec Decrease Decrease Decreased Not observed Not Not considered Increaseb
observed
Frequency of collection Increase Increase Increase Increase Increase Increase Not Increase Increase
observed
Home collection Increase Increase Increase Not considered Not considered Increase Increase Increase Decreasee
Contract out Not observed Decreasef Decrease Decrease Not observed Decrease Increase Not observed Not observed
Separate collection of Not considered Not Not considered Not considered Decreaseg Not Not Not observed Increaseh
recyclables considered considered considered
Non-domestic waste Not considered Decrease Not considered Not considered Not considered Decrease Not Not considered Decreasei
considered
a
Economies of scale are limited to small municipalities.
b
Both economies of scale and density are observed for heavy multi-material only.
c
Economies of density are limited to large municipalities.
d
Economies of density are found for waste management only.
e
Home collection of differentiated waste lowers disposal costs.
f
Privatization benefit is found for large municipalities only.
g
Municipalities can decrease waste management costs by offering joint disposal and recycling services.
h
The overall percentage of recycled waste is positively associated with paper, paperboard and organic collection costs.
i
The larger share of non-home organic waste lowers disposal costs.
 R. Chifari et al. / Waste Management 60 (2017) 32–41
tion cost. In spite of this outcome, an overall regression analysis
that evaluates determinants of MSW management costs has not
yet been conducted. Furthermore, the previously quoted study by
Kinnaman et al. (2014) on Japanese municipalities was limited to
an examination of 81 large centers.
In regards, the role of different management stages, i.e., collec-
tion, processing, and disposal, Hogg (2002) compared costs among
EU countries, reporting that the costs of residual collection range
from 42 Euro/ton (UK) to 126 Euro/ton (Denmark). The World
Bank (2012) performed an even broader study by comparing costs
incurred in developed and developing countries. According to esti-
mations made in this study, collection costs range from 20–50 US
$/ton (low income countries) to 85–250 US$/ton (high income
countries) while for the disposal stage, costs of disposal in a sani-
tary landfill range from 5–30 US$/ton (low income countries) to
35–90 US$/ton (high income countries).
In addition to the geographical aspect, past research has
also been limited in that it has rarely considered the fact that
waste management actually involves an integrated system
composed of several interacting and intertwined components
(Hogg, 2002).
In this study, we attempt to address this absence of an evalua-
tion of overall cost structures at the collection, processing, and dis-
posal stages. Specifically, we show how the separation of
recyclables affects costs at the collection, processing, and disposal
stages. Moreover, we attempt to evaluate consequent effects on the
costs of the factors listed in Table 1.
The Japanese government has applied a number of recycling
laws over the last two decades. Among them, the Containers
and Packaging Recycling Act should have had greatest effects
on household waste management (The Japan Container and
Recycling Association, 2016). Companies producing or using con-
tainers and wrapping materials have been assigned the responsi-
bility of recycling these materials in accordance with the volume
that they manufacture or sell while households are responsible
for separating these materials from solid waste and are often
requested to take them to designated collection locations.
Although Japan has achieved high recycling rates from to this
recycling law, the costs of the recycling program have not yet
been fully examined.
Waste recycling is often assessed through life-cycle assessment.
However, we focus on the financial costs of recycling in this study.
Our analysis of the financial cost structure involving related influ-
encing factors with existing financial data should serve as a prelim-
inary and relatively rapid approach to examining the cost-
effectiveness of MSW management systems (Weng and Fujiwara,
2011).
3. Data
Fig. 1 shows the trajectory (collection, processing, and disposal)
of Japanese system management stages (Fig. 1).
More specifically, the collection stage involves MSW collection
from households and transportation to intermediate treatment
facilities. At the processing stage, recyclables are recovered from
collected waste at intermediate treatment facilities, and the vol-
ume of waste is reduced at incineration facilities. Finally, at the dis-
posal stage, the processed residue is buried in a landfill.
Services delivered at different stages are often provided by dif-
ferent companies or organizations. This is why the costs of waste
management are typically estimated separately. For this reason,
obtaining an overall understanding of the full system is onerous.
For instance, while the separation of recyclable waste reduces
the number of tasks carried out at intermediate facilities, more fre-
quent collection may be required, thus increasing the workload at
the collection stage. That is, the separation of recyclable waste may
35
decrease processing costs at the expenses of raising collection
costs.
In regards to source separations, to analyze the effects of source
separation statistically, we must find an adequate dataset that
allows for a uniform standard of comparison between the waste
management costs incurred at different municipalities. While such
data are not readily available, the Japanese database perfectly ful-
fills this requirement, as all Japanese municipalities are required to
file their waste management information in a specified format and
to report these data annually to the Ministry of the Environment.
Therefore, the data collection method is standardized across all
municipalities.
The above-mentioned primary dataset was derived from the
Annual Survey of Municipal Solid Waste (Ministry of the
Environment of Japan, 2010). The survey covers all Japanese
municipalities and includes detailed information on MSW manage-
ment and recycling programs. Due to a lack of data available, we
excluded the Tokyo Special Wards and Minamisanriku in Miyagi
prefecture. The Katori municipality of the Chiba prefecture was
also excluded as its low unit cost (12.38 USD cent kg
1) renders it
an outlier. The remaining 1724 municipalities covered in the
2010 census were included in our dataset.
Some municipalities provide MSW services jointly with adja-
cent municipalities through a waste management association.
We summed each municipality’s spending on waste management
activities and financial contributions made toward waste manage-
ment associations to calculate the total management cost of MSW
services. The figure includes all operative costs for collection, pro-
cessing, and disposal activities but does not include fixed costs
such as investments for facility construction. Therefore, we focus
strictly on running costs, i.e., operations and maintenance costs.
Fig. 2 shows the distribution of unitary costs of waste management
services.
To account for differences in price levels across municipalities,
costs are adjusted based on the average taxable income of munic-
ipalities. Although the average unit cost is 33.78 US cents/kg, fares
vary considerably across municipalities. Rather, the standard devi-
ation amounts to more than 20 US cents/kg.2
According to generation sources, MSW is categorized into two
types: residential waste from individual houses and commercial
waste from larger sources such as offices, hotels, and schools.
Approximately 30.5 percent of residential waste is collected by
municipalities while the remaining is collected by waste manage-
ment companies contracted by municipalities (contracted compa-
nies). On the other hand, most commercial waste is collected by
waste management companies directly hired by firms (permitted
companies). In any case, residential and commercial waste is
mixed after collection.
The collected MSW is then transported to processing facilities to
withdraw recyclables and to reduce final disposal quantities
(Fig. 1). Households and firms are allowed to bring their waste
directly to processing facilities. While households are asked to
bring their garbage to a collection site on specified days, Japanese
houses are small and rarely have much storage space. Therefore, it
is a common practice to bring garbage directly to intermediate
facilities. In 2010, shares of residential waste, commercial waste,
and waste brought directly to processing facilities (direct waste)
amounted to 66.2, 24.2, and 9.6%, respectively.
The annual survey also includes information on municipal
spending at the collection, processing, and disposal stages. How-
ever, only total spending is reported by waste management associ-
ations. In 2010, 538 municipalities did not organize through
associations. Based on data for these non-organized municipalities,
2
Throughout the paper, we assume that 1 yen = 1.2 US cents.
36 R. Chifari et al. / Waste Management 60 (2017) 32–41

Fig. 1. Flow of MSW management services in Japan.

700
Recyclable waste
Number of municipalities

600 15.7%
500 Other waste
0.3%
400
Bulky waste
300 1.6%
200
Mixed waste
100 7.0%
Combustible
0
Non-combustible waste
-10 -20 -30 -40 -50 -60 -70 -80 -90 -100 >100
waste 70.8%
Unitary cost of waste treatment (USD cent/kg)
4.6%
Fig. 2. Distribution of unitary costs of MSW treatment (N = 1724).

the average unit costs of collection, treatment, and disposal

amount to 10.96 cent/kg, 17.28 cent/kg, and 3.08 cent/kg,
Fig. 3. Percentages of waste collected in the six sorting categories (2010).
respectively.3

According to the survey, MSW is separated into six sorting cat-

3
The average total cost incurred in these municipalities is 35.27 US cents/kg, which
egories: (1) combustible waste, (2) non-combustible waste, (3)
is slightly higher than the average total cost of 33.80 US cents/kg incurred across all mixed waste, (4) bulky waste, (5) other waste, and (6) recyclable
municipalities. waste. As is shown in Fig. 3, waste collection percentages for the
 R. Chifari et al. / Waste Management 60 (2017) 32–41 37

Table 2
Descriptive statistics.

Variable Unit Obs. Mean or share Std. Dev. Min Max

a
Waste volume (ton) 1724 22845.29 61678.94 49.00 1188540.00
Total treatment costa (thousands of dollars) 1724 56.59 138.61 0.13 2550.00
Spending on waste collectiona (thousands of dollars) 1724 20.60 68.58 0.00 1470.00
Spending on intermediate treatmenta (thousands of dollars) 1724 18.02 49.55 0.00 813.32
Spending on final disposala (thousands of dollars) 1724 2.53 11.74 0.00 410.85
Unit cost of total waste treatmenta (cent/kg) 1724 33.78 20.64 4.22 373.04
Unit cost of total waste treatmenta,d (cent/kg) 538 35.27 21.68 6.07 278.46
Unit cost of waste collectiona,d (cent/kg) 538 10.96 8.66 0.00 94.30
Unit cost of intermediate treatmenta,d (cent/kg) 538 17.28 13.34 0.00 115.62
Unit cost of final disposala,d (cent/kg) 538 3.08 5.35 0.00 50.27
Population densityb (person/km2) 1724 864.11 1776.81 2.00 14020.00
Collection sparsitya,b (1/km) 1724 11.61 5.69 0.92 63.15
Share of collection by associationa (share: 0.0–1.0) 1724 0.40 0.36 0.00 1.00
Share of collection by contracted companiesa (share: 0.0–1.0) 1724 0.82 0.33 0.00 1.00
Door-to-door collectiona (dummy) 1724 0.08
Combined collectiona (dummy) 1724 0.12
Recycling ratea (share: 0.0–1.0) 1724 19.83 9.18 0.00 80.73
Share of mixed wastea (share: 0.0–1.0) 1724 0.02 0.11 0.00 0.92
Share of combustible wastea (share: 0.0–1.0) 1724 0.73 0.18 0.00 1.00
Share of non-combustible wastea (share: 0.0–1.0) 1724 0.05 0.06 0.00 0.79
Variable Unit Obs. Mean or Share Std. Dev. Min Max
Share of bulky wastea (share: 0.0–1.0) 1724 0.02 0.03 0.00 0.31
Share of other wastea (share: 0.0–1.0) 1724 0.00 0.02 0.00 0.28
Share of recyclable waste (share: 0.0–1.0) 1724 0.18 0.12 0.00 0.96
Share of commercial wastea (share: 0.0–1.0) 1724 0.16 0.12 0.00 0.84
Share of direct wastea (share: 0.0–1.0) 1724 0.12 0.11 0.00 0.84
Altitudec (m) 1724 104.81 176.18
1.60 1186.00
Island (dummy) 1724 0.07
a
Data source: Ministry of the Environment of Japan (2010).
b
Data source: Statistics Bureau of Japan (2010).
c
Data source: Geospatial Information Authority of Japan (2016).
d
Only municipalities that do not organize any association are considered.

six sorting categories are (1) 70.8, (2) 4.6, (3) 7.0, (4) 1.6, (5) 0.3,
and (6) 15.7%, respectively. Kitchen waste, scrap paper, sanitary
products, unrecyclable plastics,4 rubber and leather items, and
clothing are collected as combustible waste. By contrast, metallic
products, ceramics, glass, and small appliances are collected as
non-combustible waste. However, 58 municipalities do not separate
combustible and non-combustible waste and simply collect
unsorted waste as mixed waste. In consideration of the variety of
recyclables managed across municipalities, newspapers, magazines,
cardboard boxes, bottles, cans, and plastic PET bottles are generally
collected as recyclable resources.
The influence of waste volumes on costs at the collection, pro-
cessing, and disposal stages is evaluated using the empirical model
presented below. In addition, the effects of compositions are esti-
mated by addressing how the six types of fractions affect MSW
management costs. The data source and descriptive statistics on
the variables addressed in the study are presented in Table 2.
4. Research methods
The empirical model was developed by modifying the standard
cost function used in the studies listed in Table 1. To examine how
the quantity and separation of waste affect the costs of waste man-
agement, we converted the quantity of total waste reported in the
survey Q T to the quantity of waste adjusted for the costs of treat-
ment Q R (all variables are presented in the legend included in
Nomenclature):
! k¼2
Y
m
QR ¼ ðsk Þhk Q T
k¼1
where sk is the share of the kth category of waste and where hk is
the parameter that measures the cost of treatment. For normaliza-
tion purposes, we set h1 ¼ 0 for combustible waste. When the kth
waste category requires more (less) intensive treatment than com-
bustible waste, hk is estimated to be larger (smaller) than 0. For
instance, suppose that s1 and s2 are the shares of combustible and
recyclable waste, respectively. When recyclable waste requires less
intensive treatment (i.e., less intensive recyclable separation at
intermediate facilities) than combustible waste, we estimate
h2 < h1 ¼ 0 at the processing stage.
Following Stevens (1978), Szymanski and Wilkins (1993) and
Tsilemou and Panagiotakopoulos (2006), we assume a Cobb-
Douglas production technology:
Q R ¼ ALa K b ð2Þ
where A is a constant that is assumed to be the same across all
municipalities; L is the total quantity of labor inputs; and K is the
total quantity of capital inputs. Parameters a and b are the shares
of output attributable to labor and capital, respectively. The cost
to the municipality is:
C ¼ wL þ rK ð3Þ
where w is the wage rate and r is the price of capital. By minimizing
(3) subject to (2) and by inserting Eq. (1), we have a log linear cost
function:
X
m
lnC ¼ c0 þ c1 lnw þ c2 lnQ T þ h0k lnsk ð4Þ
ð1Þ where c0 is the constant term that includes the price of capital,
c1 ¼ a=ða þ bÞ, c2 ¼ 1=ða þ bÞ, h0k ¼ hk =ða þ bÞ.

To better focus on the cost structure of residential waste, we

4
These exclude polyethylene terephthalate (PET) bottles and plastic packaging add the vector of control variables of X to Eq. (4) and estimate
waste, which are ‘‘recyclable.” Eq. (5),
38 R. Chifari et al. / Waste Management 60 (2017) 32–41
X
m
lnC ¼ c0 þ c2 lnQ T þ BX þ h0k lnsk :
k¼2
Eq. (5) was estimated for the three treatment stages. Parameter c2 is
the cost elasticity with respect to total-waste throughput Q T . Based
on the size of this parameter, it is possible to determine whether an
economies-of-scale effect exists or not. Parameter hk evaluates the
composition effect where a positive hk denotes that the cost of
MSW management increases as the share of kth waste increases.
Finally, marginal effects of the control variables X are evaluated
with the vector of the parameters B.
As our control variable, we use the share of commercial waste
as well as the share of waste directly brought to intermediate facil-
ities. Municipalities can save on collection costs when waste is
brought directly into intermediate facilities or is collected by pri-
vate companies directly hired by firms.
Furthermore, population density was included as a control vari-
able to examine the effects of economies of density. A collection
sparsity variable allows one to account for the road accessibility
of municipalities. This collection sparsity variable is calculated by
dividing the total road length by the residential area. The larger
this variable is, the sparser the municipality is.
Moreover, two additional dummy variables were considered to
account for service quality levels. The door-to-door collection
dummy variable takes a value of one when waste is collected from
individual households. The combined collection dummy variable
takes a value of one when door-to-door and street collection5
methods are used simultaneously.
To examine whether contracting out lowers the costs of MSW
management, we considered the share of residential waste col-
lected by contracted companies. We also considered the share of
residential waste collected by waste management associations.
When economies of scale exist in MSW management schemes, a
municipality can reduce costs by providing waste management
services in cooperation with adjacent municipalities. In addition,
monitoring from multiple parties can improve the performance
of waste management services.
Finally, two geographical variables linked to physical character-
istics of the territory studied were also considered. The island
dummy variable takes a value of one when a municipality is
located on an island or includes an island within its boundaries.
The altitude variable also allows one to consider municipalities
in mountainous areas. While traffic is normally less congested in
mountainous areas, trucks must travel up and down mountain
slopes to transport waste. Hence, the effects of altitude on manage-
ment costs are unknown a priori.
5. Empirical results
Fig. 4 describes our research flow.
In Section 5.1, we examine how the total cost of waste manage-
ment C T changes with the total waste volume Q T . We also identify
determinants of waste management costs and compare them with
those found through previous studies. In Section 5.2, we examine
how costs incurred at three treatment stages (the collection cost
C C , processing cost C P , and disposal cost C D ) change with the total
waste volume Q T . We report the cost elasticities incurred at three
treatment stages. Finally, in Section 5.3, we evaluate the composi-
tion effect to determine whether household separation affects
waste management costs. Specifically, we examine how the rela-
tive share of six types of waste sk affects waste management costs.
5
Street collection in Japan is based on a curbside-collection scheme but does not
depend on the use of a waste bin.
5.1. Total cost
ð5Þ
The results of our empirical model are presented in Table 3.
Eq. (5) was initially regressed without including waste-fraction
share variables, i.e., the composition of waste throughputs was not
taken into account at this stage. The second column presents the
estimation results of the total cost model. The estimated coefficient
for the log of total waste is 0.89, meaning that a one percent
increase in the total-waste throughput leads to a 0.89% increase
in the total cost. Hence, the result reveals the presence of an
economy-of-scale effect. We also analyzed the sub-samples of
large municipalities with populations exceeding 50,000 inhabi-
tants and confirm the existence of economies of scale.
The coefficient for the log of population density is
0.07. A neg-
ative coefficient denotes that there are economies of density, i.e., a
one percent increase in population density leads to a 0.07%
decrease in the total cost. In regards the collection sparsity vari-
able, the coefficient is positive and significant. This result clearly
suggests that the costs are larger in an area that is difficult to
access by vehicle.
Furthermore, Table 3 shows that costs are dependent on the
quality of services delivered. Both door-to-door and combination-
collection variables become positive and significant. In addition,
the coefficient of the door-to-door collection variable is larger than
that of combined collection. These results imply that door-to-door
collection is more costly than street collection. According to our
estimations, door-to-door collection services increase costs by
45.12%.6
A positive sign for the recycle rate denotes that recycling
increases the overall costs of MSW management. According to
our estimations, a one percent increase in the recycling rate should
cause a 0.4% increase in the total costs of MSW management in the
median municipality.
The table also shows that market structures determine MSW
management costs. According to our calculations, contracting to
external firms lowers costs by 10.6%. However, the level of cost
savings achieved in Japan is less than that found for the UK
(Szymanski and Wilkins, 1993). Another finding shows that collec-
tion by waste management organizations lowers costs by 23.0%.
Together with the fact that MSW management results in IRS during
production, we argue that the size of Japanese municipalities is less
impactful than the cost-minimizing level.
Finally, geographical variables affect MSW management costs.
Costs are higher when a municipality is located on an island or
includes an island within its boundaries (positive coefficient). This
is not surprising given the typical scarcity of land on islands and as
shipments to the mainland may be required in such geographic
contexts. In contrast, costs are lower for municipalities in high-
altitude regions, and therefore fuel savings from low levels of traf-
fic congestions outpace the more significant requirements of col-
lection routes presenting height gradients.
5.2. Costs of Single waste management stages
Waste-management associations can be organized in several
different ways. Some municipalities typically coordinate all waste
management services from collection to disposal with adjacent
municipalities while others only coordinate on processing and dis-
posal services. Other municipalities jointly use an incineration
facility but do not share resource recovery facilities. Unfortunately,
the MSW survey for Japan does not provide data on how munici-
6
The door-to-door collection variable is a dummy variable and the percentage
effect on cost can be measured by 100  eðb
1Þ where b is the coefficient of the door-
to-door dummy variable. Using this formula, we can also estimate the effects of
contracting out and organization collection.
 R. Chifari et al. / Waste Management 60 (2017) 32–41 39

Fig. 4. Research flow and empirical models.

Table 3
Cost structure of MSW management.

Total sample Sub sample (non-organized municipalities)

Total Total Collection Processing Disposal
Constant 4.93*** 4.83*** 3.02*** 4.00*** 3.55***
Log of total waste 0.89*** 0.87*** 0.96*** 0.84*** 0.73***
Log of population density
0.07***
0.04***
0.08***
0.02
0.06
Collection sparsity 0.010*** 0.008** 0.002 0.026***
0.005
Door-to-door collection 0.20*** 0.10* 0.25** 0.00 0.08
Combination collection 0.12*** 0.07 0.13*
0.02 0.00
Recycle rate 0.004*** 0.003 0.011***
0.004
0.004
Share of contracted company
0.11***
0.06
0.11 0.04
0.07
Share of association
0.26***
Share of business waste
0.60***
0.18
0.50 0.15 0.34
Share of direct waste
0.64***
0.54***
0.82***
0.38
0.08
Island 0.11** 0.16** 0.14 0.24**
0.01
Altitude
1.60E
04**
1.66E
05 2.25E
04 7.37E
05 1.09E
04
Samplea 1734 538 533 524 510
R-squared 0.93 0.95 0.88 0.82 0.50
a
Municipalities without corresponding cost data were removed from the dataset.
*
Denote statistical significance at the 10% levels, respectively.
**
Denote statistical significance at the 5% levels, respectively.
***
Denote statistical significance at the 1% levels, respectively.

palities coordinate waste management services. By focusing on
non-organized municipalities (referred to as a ‘‘sub sample”), we
analyze the cost structures of collection, processing, and disposal
stages separately. The number of municipalities included in the
sub sample was reduced to approximately one-third of the entire
sample as stated in the previous section.
The right side of Table 3 shows the results estimated at each
stage. Some variables in the sub sample models became statisti-
cally insignificant due to a reduction in the sample size. However,
we found that major variables still affect the waste management
cost. The elasticities of costs with respect to waste throughput at
the collection, processing, and disposal stages are 0.96, 0.84, and
0.73, respectively. Hence, although economies of scale are visible
at all three stages, the collection cost is less elastic than the other
two costs.
The door-to-door and combination-collection variables only
become statistically significant at the collection stage. Similarly,
the share of direct waste and the log of population density only
become significant at the collection stage. Thus, the collection
method does not affect processing and disposal costs.
A similar empirical result was obtained for the recycle rate vari-
able. In Japan, municipalities incur the collection costs of packag-
ing waste while recycling costs are incurred by producers. This
explains why the recycle rate variable is only significant for collec-
tion costs.
The results obtained for collection sparsity and island variables
also present an interesting trend. The variables are statistically sig-
nificant at the processing stage but not at the collection stage. This
can be attributed to the fact that sparsely distributed municipali-
ties tend to use small processing facilities. As the unit costs of
40 R. Chifari et al. / Waste Management 60 (2017) 32–41
Table 4
Effects of waste compositions on total, separation, and disposal costs (Combustible waste is used as a base.)
Sub sample (non-organized municipalities)a
Total
Share of mixed waste
0.259*
Share of non-combustible waste
0.561*
Share of bulky waste 0.387
Share of other waste
0.209
Share of recyclable waste
0.105
Sample 538
a
Municipalities without corresponding cost data were removed from the dataset.
*
denote statistical significance at the 10% levels, respectively.
**
denote statistical significance at the 5% levels, respectively.
***
denote statistical significance at the 1% levels, respectively.
running such small facilities are generally higher, the collection
sparsity variable only becomes statistically significant at the pro-
cessing stage. For islands, MSW generated must be transported
to a processing facility by ship. The logistical cost of marine trans-
portation is included in the overall commission fee at the process-
ing stage while fees for MSW collection within the island are paid
by the municipality. Consequently, the ‘‘island” variable is only sta-
tistically significant at the processing stage.
5.3. Waste separation effects on costs
The coefficients obtained by introducing the shares of various
waste fractions are presented in Table 4. Only a sub-set of data is
presented, as the results of the remaining variables are similar to
those shown in Table 3. As stated above, combustible waste is used
as a base in this analysis. Table 4 shows that the composition of
waste affects costs at the three stages in different ways.
The negative sign at the processing stage suggests that com-
bustible waste requires more intensive treatment than other types
of waste at intermediate treatment facilities. This could be attrib-
uted to the greater labor requirements needed to separate recy-
clables at recycling facilities or to the greater energy needs of
incineration facilities. On the other hand, the positive sign at the
disposal stage suggests that combustible waste requires less
expensive treatment at final disposal sites and likely due to the
reduction in waste to be disposed of after incineration.
Furthermore, Table 4 shows that recyclable waste reduces costs
at the processing stage but does not change the total cost of MSW
management. Thus, the separation of recyclables from non-
recyclables may improve efficiency of intermediate treatment
facilities but should not improve the efficiency of overall waste
management services.
6. Conclusion and policy implications
In this paper, we elaborated on a cost structure analysis of MSW
management that considers two factors that have been overlooked
in the literature. First, cost elasticity was estimated in relation to
waste volume at system management stages (collection, process-
ing, and disposal). The effects of other variables such as market
structures and geographic contexts were accounted for. Amidst
these factors, we found economies of scale at all three stages;
moreover, we found that municipalities can reduce waste manage-
ment service costs by 26% when a waste management association
forms among adjacent municipalities. This cost reduction is much
more significant than that achievable through privatization, which
would amount to only 11%.
The Ministry of the Environment of Japan (1997) introduced
guidelines of regional waste management programs referred to
as the Area Wide Program of Waste Disposal. Since then, Japan
Collection Processing Disposal
0.496***
1.661*** 0.806**
0.374
2.229*** 3.607***
1.761
2.146 2.627
1.839***
0.026
0.594
0.453
0.901** 0.821
533 524 510
has experienced the great Heisei merger and the number of munic-
ipalities decreased from 3229 in 1999 to 1727 in 2010. Despite this
large-scale merger, this study based on 2010 data shows that there
is still further scope for cost-savings through regional waste man-
agement programs (e.g., by consolidating collective processing
facilities, it is possible to achieve better cost- and energy
recovery-efficient management).
Our comparisons of estimated parameters across the three
waste management stages suggest that disposal costs are more
elastic than collection costs. Therefore, when waste volumes are
reduced, disposal costs should most likely decrease more than col-
lection costs. More specifically, a one percent increase in the total
waste volume lowers the collection cost per unit by 4% and the dis-
posal cost per unit by 27%.
In regards to waste separation, costs incurred at the three waste
management stages are affected in different ways. For instance,
combustible waste lowers collection and disposal costs but
increases processing costs. On the other hand, recyclable waste
lowers processing costs but does not change the overall manage-
ment cost.7
Although we focus on the economic aspects of MSW manage-
ment in this paper, it is also important to determine how environ-
mental and social conditions affect the performance of MSW
management schemes (De Feo and De Gisi, 2014; Ferreira et al.,
2014; Münster et al., 2015). For instance, assessing the costs of
air pollution through incineration (Brunner and Rechberger,
2015; Eriksson and Bisaillon, 2011) and evaluating the effects of
waste collection schemes on local employment patterns would
expand the scope of this analysis. Indeed, municipal waste man-
agement system performance can be only evaluated by acknowl-
edging the presence of several incommensurable and potentially
contrasting dimensions (Chifari et al., 2016; Blengini et al., 2012).
Sources of uncertainty (Cai et al., 2007; Clavreul et al., 2012; Dai
et al., 2014), data quality levels, and global sensitivities must also
be considered in such assessments. To this end, a participatory pro-
cess must be used to determine the criteria most relevant to local
stakeholders. These factors, complementary to our evaluation of
financial dimensions of MSW, will be examined in future studies.
Acknowledgements
The corresponding author received financial support from the
Institute of Economic Research of Aoyama Gakuin University and
from the Japan Society for the Promotion of Science KAKENHI
(Grant Number: 26340119). Much of this study was conducted
over his sabbatical year at the Institute for Environmental Sciences
and Technology, Universitat Autònoma de Barcelona. The authors
7
A more thorough discussion would require taking into account financial benefits
derived from material and energy recovery from waste.
 R. Chifari et al. / Waste Management 60 (2017) 32–41
wish to thank Mario Giampietro and Sandra Bukkens for coordinat-
ing our research. An earlier version of this paper was presented at
the annual conference of the Society for Environmental Economics
and Policy Studies, the 7th International Research Meeting in Busi-
ness and Management, and the 15th International Conference of
the Japan Economic Policy Association. We acknowledge com-
ments and suggestions received from the conference participants.
References
Bari, Q.H., Mahbub Hassan, K., Haque, R., 2012. Scenario of solid waste reuse in
Khulna city of Bangladesh. Waste Manage. 32, 2526–2534.
Bel, G., Fageda, X., 2010. Empirical analysis of solid management waste costs: some
evidence from Galicia, Spain. Resour. Conserv. Recy. 54, 187–193.
Blengini, G.A., Fantoni, M., Busto, M., Genon, G., Zanetti, M.C., 2012. Participatory
approach, acceptability and transparency of waste management LCAs: Case
studies of Torino and Cuneo. Waste Manage. 32, 1712–1721.
Brunner, P.H., Rechberger, H., 2015. Waste to energy - key element for sustainable
waste management. Waste Manage. 37, 3–12.
Bosch, N., Predaja, F., Suárez-Pandiello, J., 2000. Measuring the efficiency of Spanish
municipal refuse collection services. Local Gov. Stud. 26, 71–90.
Cai, Y., Huang, G.H., Nie, X.H., Li, Y.P., Tan, Q., 2007. Municipal solid waste
management under uncertainty: a mixed interval parameter fuzzy-stochastic
robust programming approach. Environ. Eng. Sci. 24, 338–352.
Callan, S.J., Thomas, J.M., 2001. Economies of scale and scope: a cost analysis of
municipal solid waste services. Land Econ. 77, 548–560.
Castagna, A., Casagranda, M., Zeni, A., Girelli, E., Rada, E.C., Ragazzi, M., Apostol, T.,
2013. 3R’S from citizens point of view and their proposal from a case-study. UPB
Sci. Bull. Ser. D Mech. Eng. 75, 253–264.
Chifari, R., Lo Piano, S., Bukkens, S.G.F., Giampietro, M., 2016. A holistic framework
for the integrated assessment of urban waste management systems. Ecol. Indic.
Clavreul, J., Guyonnet, D., Christensen, T.H., 2012. Quantifying uncertainty in LCA-
modelling of waste management systems. Waste Manage. 32, 2482–2495.
Collins, J.N., Downes, B.T., 1977. The effect of size on provision of public services.
The case of solid waste collection in smaller cities. Urban Affairs. Q. 12, 333–
347.
Dai, C., Cai, X.H., Cai, Y.P., Huo, Q., Lv, Y., Huang, G.H., 2014. An interval-parameter
mean-CVaR two-stage stochastic programming approach for waste
management under uncertainty. Stoch. Environ. Res. Risk Assess. 28, 167–187.
De Feo, G., De Gisi, S., 2014. Using MCDA and GIS for hazardous waste landfill siting
considering land scarcity for waste disposal. Waste Manage. 34, 2225–2238.
De Jaeger, S., Eyckmans, J., 2015. From pay-per-bag to pay-per-kg: The case of
Flanders revisited. Waste Manage. Res. 33, 1103–1111.
Dijkgraaf, E., Gradus, R.H.J.M., 2003. Cost savings of contracting out refuse
collection. Empirica 30, 149–161.
Di Leo, S., Salvia, M., 2017. Local strategies and action plans towards resource
efficiency in South East Europe. Renew. Sustain. Energy Rev. 68, 286–305.
Domberger, S., Meadowcroft, S.A., Thompson, D.J., 1986. Competitive tendering and
efficiency: the case of refuse collection. Fisc. Stud. 7, 69–87.
Dubin, J.A., Navarro, P., 1988. How markets for impure public goods organize the
case of household refuse collection. J Law Econ Organ 4, 217–241.
Eriksson, O., Bisaillon, M., 2011. Multiple system modelling of waste management.
Waste Manage. 31, 2620–2630.
Ferreira, S., Cabral, M., da Cruz, N.F., Simões, P., Marques, R.C., 2014. Life cycle
assessment of a packaging waste recycling system in Portugal. Waste Manage.
34, 1725–1735.
Greco, G., Allegrini, M., Del Lungo, C., Savelliniet, P.G., Gabellini, L., 2015. Drivers of
solid waste collection costs. Empirical evidence from Italy. J. Clean. Prod. 106,
364–371.
Hirsch, W.Z., 1965. Cost functions of an urban government service: refuse
collection. Rev. Econ. Stat. 47, 87–92.
Kinnaman, T., 2006. Policy watch: examining the justification for residential
recycling. J. Econ. Perspect. 20, 219–232.
Kinnaman, T., Shinkuma, T., Yamomoto, M., 2014. The socially optimal recycling
rate: evidence from Japan. J. Environ. Econ. Manage. 68, 54–70.
Kitchen, H.M., 1976. A statistical estimation of an operating cost function for
municipal refuse collection. Public. Financ. Quart. 4, 56–76.
41
Massarutto, A., de Carli, A., Graffi, M., 2011. Material and energy recovery in
integrated waste management systems: a life-cycle costing approach. Waste
Manage. 31, 2102–2111.
Münster, M., Ravn, H., Hedegaard, K., Juul, N., Ljunggren Söderman, M., 2015.
Economic and environmental optimization of waste treatment. Waste Manage.
38, 486–495.
Ohlsson, H., 2003. Ownership and production costs. Choosing between public
production and contracting-out in the case of Swedish refuse collection. Fisc.
Stud. 24, 451–476.
Reeves, E., Barrow, M., 2000. The impact of contracting-out on the costs of refuse
collection services. The case of Ireland. Econ. Soc. Rev. 31, 129–150.
Ren, X., Hu, S., 2014. Cost recovery of municipal solid waste management in small
cities of inland China. Waste Manage. Res. 32, 340–347.
Stevens, B.J., 1978. Scale, market structure, and the cost of refuse collection. Rev.
Econ. Statist. 60, 438–448.
Szymanski, S., 1996. The impact of compulsory competitive tendering on refuse
collection services. Fisc. Stud. 17, 1–19.
Szymanski, S., Wilkins, S., 1993. Cheap rubbish? Competitive tendering and
contracting out in refuse collection. Fisc. Stud. 14, 109–130.
Tickner, G., McDavid, J.C., 1986. Effects of scale and market structure on the costs of
residential solid waste collection in Canadian cities. Public Finan. Quart 14 (4),
371–393.
Tsilemou, K., Panagiotakopoulos, D., 2006. Approximate cost functions for solid
waste treatment facilities. Waste Manage. Res. 24, 310–322.
Saeed, M.O., Hassan, M.N., Mujeebu, M.A., 2009. Assessment of municipal solid
waste generation and recyclable materials potential in Kuala Lumpur, Malaysia.
Waste Manage. 29, 2209–2213.
Sarkhel, P., Banerjee, S., Banerjee, S., 2016. Willingness to pay before and after
program implementation: the case of Municipal Solid Waste Management in
Bally Municipality, India. Environ. Dev. Sustain. 18, 481–498.
Troschinetz, A.M., Mihelcic, J.R., 2009. Sustainable recycling of municipal solid
waste in developing countries. Waste Manage. 29, 915–923.
Weng, Y.-C., Fujiwara, T., 2011. Examining the effectiveness of municipal solid
waste management systems: an integrated cost–benefit analysis perspective
with a financial cost modeling in Taiwan. Waste Manage. 31, 1393–1406.
Zabaleta, I., Rodic, L., 2015. Recovery of essential nutrients from municipal solid
waste – impact of waste management infrastructure and governance aspects.
Waste Manage. 44, 178–187.
Zhu, J., Huang, W., Sun, W., Huang, G., 2016. Waste management model associated
with public-private partnership in Hamilton, Ontario, Canada. J. Environ. Eng.
142.
Web references
Geospatial Information Authority of Japan, 2016. National Land Numerical
Information Download Service. <http://nlftp.mlit.go.jp/ksj-e/index.html>
(access 20.05.16).
Hogg, D., 2002. Costs for municipal waste management in the EU. Final Report to
Directorate General Environment, European Commission. <http://ec.europa.eu/
environment/waste/studies/pdf/eucostwaste.pdf> (access 10.08.16).
Ministry of the Environment of Japan, 1997. Area Wide Program of Waste Disposal.
<http://www.env.go.jp/hourei/11/000138.html> (access 26.02.16).
Ministry of the Environment of Japan, 2010. The Annual Survey of Municipal Solid
Waste. <http://www.env.go.jp/recycle/waste_tech/ippan/> (access 20.02.16).
Statistics Bureau of Japan, 2010. Statistical observations of shi, ku, machi, mura.
<http://www.stat.go.jp/data/s-sugata/> (access 05.02.16).
The Japan Container and Recycling Association, 2016. Point of act: What is the
Containers and Packaging Recycling Act? <http://www.jcpra.or.jp/english/
tabid/614/index.php> (access 09.12.16).
World Bank, 2011. Urban solid waste management. <http://web.worldbank.org/
WBSITE/EXTERNAL/TOPICS/EXTURBANDEVELOPMENT/EXTUSWM/0,,menuPK:
463847~pagePK:149018~piPK:149093~theSitePK:463841,00.html> (access
26.02.16).
World Bank, 2012. What a waste: A global review of solid waste management.
<http://siteresources.worldbank.org/INTURBANDEVELOPMENT/Resources/
336387-1334852610766/What_a_Waste2012_Final.pdf> (access 26.02.16).