Journal of Environmental Management 214 (2018) 416e425

Contents lists available at ScienceDirect

Journal of Environmental Management

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

Research article

Towards sustainable consumption: A socio-economic analysis of

household waste recycling outcomes in Hong Kong
Alex Y. Lo a, *, Shuwen Liu b
a
Department of Geography, University of Hong Kong, 10/F, The Jockey Club Tower, Pokfulam Road, Hong Kong
b
School of Arts and Social Sciences, The Open University of Hong Kong, 7/F, Block A, 30 Good Shepherd Street, Homantin, Kowloon, Hong Kong

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

Article history: Many high-density cities struggle to find space for disposing municipal solid waste. Hong Kong is one of
Received 23 August 2017 these cities, seeking to scale up waste recovery efforts as an alternative to disposal. However, territory-
Received in revised form wide recovery initiatives do not account for socio-economic variations across place, leading to mixed
5 February 2018
outcomes among diverse communities. This study aims to investigate socio-economic effects on recy-
Accepted 6 March 2018
Available online 13 March 2018
cling behavior in a sample of subsidized rental housing estates. It constitutes an improvement from
previous studies by using the entire estate as a unit of analysis and analyzing actual recycling outcomes,
which have received limited attention from researchers. The analysis focused on the volume of re-
Keywords:
Household waste
cyclables collected from 158 public housing estates in Hong Kong, with an average population of 12,285.
Waste separation Results suggest that recycling outcomes vary with a limited set of socio-economic factors. Housing es-
Waste recycling tates managed by a private property management company and populated by better off households
Environmental behavior collected more recyclables from their residents. Measures of absolute and relative recycling intensity
Household income achieved similar results. The findings will be useful for identifying residential communities requiring
additional support for promoting waste separation and recycling. Differentiated policies for economically
disadvantaged communities are warranted.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction favor the use of centralized household waste recycling facilities

High-rise and high-density development mode is prevalent in
Hong Kong, a Special Administrative Region (SAR) of China. An
excess of 7.3 million people inhabit a tiny share of land - residential
areas across the city are restricted to approximately 77 km2, or 6.9
per cent of total land area (Planning Department, 2015). The city is
divided into 18 Districts, each overseen by a District Council. Some
of the urban Districts have been filled to over 45,000 persons/km2.
In the built-up area, the majority of residential developments are
estate-based, in which housing blocks are developed by the same
developer(s) and managed by the same property management
agent (Yau, 2010). A typical housing block is 20e40 stories high and
has hundreds of self-contained apartments. Large residential es-
tates consist of dozens of housing blocks. The five largest ones in
Hong Kong are populated by more than 36,000 people each (Census
and Statistics Department, 2012).
The high-rise and high-density developments in Hong Kong
* Corresponding author.
E-mail address: alexloyh@hku.hk (A.Y. Lo).
(Chung and Poon, 1994). The most widely used of these are the
tricolor recycling bins (or waste separation bins) placed at
numerous locations across the city, including thousands of resi-
dential sites (Fig. 1). According to the Environment Bureau (2013),
over 80 per cent of Hong Kong residents can find recycling bins in
close proximity. They can separate out metal cans, plastic products,
and newsprint/magazines and paper products (such as packaging)
and place them in the proper bins close to their home. Access to
these facilities is usually provided at communal areas, such as the
ground floor of multi-family housing blocks, podiums, central car-
parks, and other open spaces within gated residential communities.
Recycling outcomes crucially hinge upon householders' partic-
ipation (Ko and Poon, 2009; Pakpour et al., 2014; Perrin and Barton,
2001; Thi et al., 2015). Many studies have sought to ascertain what
drive their participation in waste sorting and recycling (Cheung
et al., 2017; Chung and Leung, 2007; Chung and Poon, 1994;
Ekere et al., 2009; Lee and Paik, 2011; Lo, 2016b; Martin et al., 2006;
Miliute-Plepiene et al., 2016; Tucker et al., 1998). Most of them have
given consideration to the socio-economic (and demographic)
profiles of individual householders, but evidence is mixed and far

https://doi.org/10.1016/j.jenvman.2018.03.029
0301-4797/© 2018 Elsevier Ltd. All rights reserved.
 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425
Fig. 1. An example of tricolor recycling bin in a public rental housing estate in Hong Kong.
Source: Author
from consistent (Miafodzyeva and Brandt, 2013). Empirical ana-
lyses on income effects are particularly enlightening. They can help
predict how such a sustainability practice (i.e. waste recycling) will
evolve in increasingly affluent societies. Contrary to usual expec-
tations, however, it is far from clear that income or wealth drives
such practice (Brechin, 1999; Dunlap and York, 2008). The con-
tested debates warrant the use of a varied methodological strategy
in research.
Furthermore, socio-economic analysis of recycling outcomes is
highly relevant to policy-making. It can help identify particular
communities and groups of people with low participation rates,
and assess the cost-effectiveness of waste recycling schemes and
justify public spending (da Cruz et al., 2014; Simo~es and Marques,
2012). The Hong Kong SAR Government was aware, and its
Census and Statistics Department (CSD) had conducted an omnibus
household survey 16 years ago (Census and Statistics Department,
2002), which included a section on the public's participation in
source separation and recycling of domestic waste. Not only has
this survey lost currency, its methodology is bound to a self-
reporting approach. Chung and Leung (2007) and Yau (2010) have
indicated that this methodological limitation could compromise
the quality of scholarly studies on waste recycling in Hong Kong.
Martin et al. (2006, p. 359) from the UK arrived at the same
observation. Citing several UK studies, they noted that actual
participation in waste recycling often fell short of expressed
commitment. These authors call for an outcome-based approach,
which may involve an inquiry into the actual amount of recyclables
collected. Gamberini et al. (2013) and Sidique et al. (2010) have
made such a pioneering attempt to examine the relationship be-
tween the amount of waste collected from selected communities
and their characteristics (e.g. population density, per capita
income).
The present research constitutes an improvement from previous
studies by adopting a community-based and outcome-based
417
approach. The main objective is to investigate socio-economic ef-
fects on waste recycling behavior. We address questions about how
recycling outcomes vary with socio-economic traits, such as in-
come, age, gender, education, household size, across place. Such an
analysis can help identify which type of residential communities
require additional support for promoting waste separation and
recycling, and inform administrative and funding decisions of local
governments and other public institutions. It will also contribute to
the scholarly debates on socio-economic effects by using large,
subsidized rental housing estates with an average population of
12,285 as a unit of analysis. The new evidence generated will pro-
vide an indication on actual community behavior, instead of self-
reported individual behavior, creating an alternative empirical ba-
sis for re-visiting previous scholarly and official studies.
This research involves an empirical analysis of the volume of
recyclables collected from 158 public rental housing estates in
Hong Kong against socio-economic and demographic variables
derived from population census. The remainder of this paper begins
with an outline of the policy context in which the study is situated.
A brief review of literature is then presented in Section 3, followed
by a detailed explanation on the methods used in Section 4. Results
are reported and discussed in Section 5. Conclusions are elaborated
in Section 6.
2. Background
Municipal solid waste (MSW) management has been one of the
major policy challenges for the Hong Kong SAR Government. In the
past three decades, the volume of MSW has increased by nearly 80
per cent, whereas the city's population grew by 36 per cent only
(Environment Bureau, 2013). As a driver of production as well as
consumption, economic growth contributed to the increase in the
amount of waste disposed in the early years (Ko and Poon, 2009).
Domestic waste has been the biggest component (61.8%) of MSW,
418 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425

reaching 6391 tonnes per day, or 2.34 million tonnes throughout

2016 (Environmental Protection Department, 2017). Recyclable
materials placed into recycling bins and transferred to recycling
outlets are counted as recovery quantity (Environment Bureau,
2013, p. 3). In 2016, a quarter of domestic waste collected from
residential areas was recovered (Environmental Protection
Department, 2017), and the rest was transferred and disposed at
three landfill sites.
Thirty-year efforts in managing domestic waste have produced
mixed outcomes. Back in 1988, a trial on-site collection of plastic
waste at a housing estate ended up in a sense of disappointment,
because of limited participation by local residents (Chung and
Poon, 1994). The amount of domestic waste continued to increase
by an annual rate of 4 per cent during the period of 1983e1996
(Environmental Protection Department, 1997). Prohibitive space
constraints on landfill expansion prompted the Hong Kong Gov- Fig. 2. Trends of domestic waste disposal and recovery in Hong Kong (2006e2016).
ernment to review its MSW management strategies. In the second Source: Environmental Protection Department. Waste Data and Statistics - Monitoring
half of 1990s, the Government set up special task forces and of Solid Waste in Hong Kong 2006e2016 (www.wastereduction.gov.hk). The annual
launched new programs for waste reduction in order to raise the monitoring reports provide statistics on the quantity of disposed domestic waste
(absolute value) and the proportion of recovered domestic waste (percentage). Since
domestic waste recovery rate from about 8 per cent in 1997 (Chan,
generation quantity (g) is the sum of recovery quantity (r) and disposal quantity (d)
1998) to 14 per cent in 2004 (Environmental Protection (Environment Bureau, 2013, p. 3), then g ¼ d/(1er/g). As both the value of d and r/g are
Department, 2005). In 2005, the Government, under a new known, the quantity of recovered domestic waste can be calculated. Per capita figures
administration, raised the target to 26 per cent by 2012 are then calculated by dividing the total quantity over the total population of Hong
Kong estimated by the Census and Statistics Department
(Environmental Protection Department, 2005). A hallmark of its
ten-year policy framework, running from 2005 to 2014, was the
Programme on Source Separation of Domestic Waste (SSWP).
that of private housing dwellers (i.e.HK$29,610, or approx.
The Environmental Protection Department (EPD) launched the
US$3800) and all households in Hong Kong (i.e. HK$20,500, or
citywide SSWD Programme in 2005 to encourage more house-
approx. US$2630) (Census and Statistics Department, 2012). This is
holders to separate their domestic waste for recycling by providing
partly due to the fact that applicants for public rental apartments
additional waste sorting facilities and broadening the types of re-
are subject to upper limits on household income, above which they
cyclables to be recovered. It began as a pilot project in 13 housing
would be rendered ineligible. Occupants are required to vacate
estates in the East of Hong Kong Island in 2004. Broad-base ini-
their units as long as their income levels exceed these limits.
tiatives included free distribution of recycling bins to participating
Despite the socio-economic disadvantages of their residents,
housing estates or buildings, and an earmarked funding scheme
public rental housing estates have no access to the ECF's special
worth of HK$5 million (approx. US$0.64 million) administered by
funds dedicated to promoting waste recycling and separation at
the Environment and Conservation Fund (ECF) for encouraging
housing estates, for the reason that “they may receive other forms
participation by residents' organizations and property manage-
of Government funding” (Environment and Conservation Fund,
ment companies in the Programme and covering part of the
2016, p. 3). In the SSWP Programme, their observed performance
upfront costs incurred. The EPD, ECF, and District Councils also
in recycling appears falling behind their private housing counter-
operate a variety of educational and promotional schemes. With
parts.1 A closer scrutiny, presented in Section 5, confirms that
these measures and schemes, the Government promised to achieve
public housing estates achieved lower numerical results. A house-
a 50 per cent increase in the volume of waste recovery in housing
hold survey conducted by Ko and Poon (2009) also suggests that
estates after the first year of implementation and bring 80 per cent
the frequency and intensity of waste recycling and separation was
of the population in Hong Kong on board by 2010. The objective was
lower among public housing dwellers. The ECF's perverse funding
to increase the number of housing estates under the Programme by
approach and the Government's financial support to these estates
an order of magnitude, eventually reaching 1360 by the end of 2010
in this specific policy area require reconsideration in this light.
(Environmental Protection Department, 2005). The number has
Nonetheless, Chung and Poon (1994) earlier socio-economic
elevated to 2130 in December 2016.
analysis has shown that residents of public housing estates indi-
A resounding success is yet to achieve. Between 2006 and 2015,
cated stronger support to waste recycling and minimization mea-
the volume of disposed domestic waste recorded a steady increase
sures. The 2002 household survey conducted by the CSD also
in parallel to population growth. Alarmingly, however, there was a
indicated a similar tendency. This seems to be at odds with the
decline in the volume of recovered domestic waste since 2010,
actual results under the SSWP Programme. Clearer evidence is
touching down 0.28 kg per day per capita in 2016 (Fig. 2). The
needed as to what characteristics of these estates are associated
overall MSW recovery rate has also winded down, falling from 52
with better (or poorer) outcomes. This inquiry seeks to link socio-
per cent in 2010 to 34 per cent in 2016 (Environmental Protection
economic factors with recycling outcomes, and is informed by a
Department, 2015, 2016, 2017).
series of previous research discussed in the next section.
Public rental housing estates require attention. These housing
estates are one of the three major housing types in Hong Kong,
accommodating 2 million people, or 30.1 per cent of the total
population in 2011 (Table 1). They had a larger aging and working
class population than elsewhere in Hong Kong. Differences in
gender and average household size were modest. It is worth noting 1
This initial assessment is based on the Programme's 2015/16 Commendation
that public renters attracted lower incomes. Many of them made a Scheme. Higher awardees are predominantly private housing estates or buildings
subsistence living; their median monthly domestic household in- (www.wastereduction.gov.hk/en/household/source_achievements.htm). Accessed
come was HK$12,000 (approx. US$1540) in 2011, much lower than 19 June 2017.
 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425
Table 1
Basic demographics of public renters in Hong Kong (2011).
Population
Number of domestic households
Aged 65 or above
Number of persons with post-secondary qualifications
Number of female residents
Average household size
Median Monthly Domestic Household Income (HK$)
a
Include subsidized home ownership housing, private permanent housing, non-domestic housing, and temporary housing.
b
Across all housing types, including public rental housing.
Source: Census and Statistics Department, 2012
3. Waste recycling and socio-economic characteristics
According to the CSD's household survey, the percentage of
households in public rental housing having the habit of separating
recyclable waste materials before disposal was 50.8 per cent, higher
than that of private housing (47.4 per cent) by a small margin
(Census and Statistics Department, 2002). Generally consistent
with Chung and Poon (1994) report, the CSD survey also shows that
household participation in source separation and recycling of do-
mestic wastes is a function of household income, gender, age, and
educational attainment. However, international studies using
advanced statistical techniques have found a mixed picture.
The most frequently investigated socio-economic and de-
mographic variables include income, age, gender, education, and, to
a lesser extent, dwelling type. Several international studies have
suggested that recyclers tend to be mature, educated, and affluent
homeowners (Belton et al., 1994; Martin et al., 2006; Perrin and
Barton, 2001; Tucker et al., 1998; Vencatasawmy et al., 2000). In
others (Ekere et al., 2009; Miliute-Plepiene et al., 2016), most of
these variables do not demonstrate significant impacts on reported
recycling behavior. Based on a meta-analysis, Miafodzyeva and
Brandt (2013) have come to a similar conclusion and describe the
role of education and gender as ambiguous. Nonetheless, Chan
(1998) finds females and the younger ones more actively partici-
pating in recycling, whereas Chung and Poon (2001) show that
males and less educated groups are more likely to separate and
recover waste than others. Chung's later study (Chung and Leung,
2007) and some others (e.g. Lee and Paik, 2011) fail to find evi-
dence on the role of gender.
Besides age, income is believed to be a critical factor driving
recycling participation (Lee and Paik, 2011; Miafodzyeva and
Brandt, 2013). The broader concept of affluence is also recognized
(Martin et al., 2006; Perrin and Barton, 2001; Vencatasawmy et al.,
2000). Coggins (1994) suggests that socio-economic traits are
implicated in recycling awareness and attitude. It is conventionally
assumed that individuals with a lower socio-economic profile are
less environmentally aware than their wealthier counterparts. Ac-
cording to Coggins (1994), these individuals tend to put less effort
in recycling, as they are preoccupied with more pressing needs (e.g.
economic security) than environmental protection. Such an
explanation is derived from contemporary thoughts on social
change.
The positive income or wealth effects are a well-debated issue in
the study of environmental concern and behavior. A classic and
influential position, most notably advocated by Ronald Inglehart
(1990), is that wealthier societies are more environmentally
conscious and active than those at a lower level of development. As
societies develop and accumulate wealth, people's values have
gradually shifted towards post-materialism. This results in stronger
419
Public rental housing All other housing typesa
Estimate Proportion (%) Estimate Proportion (%)
2,074,578 30.1 4,817,081 69.9
720,892 30.4 1,647,904 69.6
331,957 16.0 528,726 11.0
249,080 12.0 1,414,578 29.4
1,075,386 51.8 3,677,078 53.4
2.9 e 2.9b e
12,000 e 20,500b e
public support for environmental protection and sustainable con-
sumption (Franzen and Meyer, 2010; Franzen and Vogl, 2013). The
post-materialist hypothesis, however, is highly contested. Some
researchers have identified a different tendency, i.e. people from
less advanced societies have indicated higher levels of willingness
to make economic sacrifices for the environment and concern
about the risks arising from environmental degradation (Brechin,
1999; Dunlap and York, 2008; Gelissen, 2007; Kim and Wolinsky-
Nahmias, 2014; Lima et al., 2005; Lo, 2014). The debate remains
unresolved as to how income or wealth influences environmental
concern and behavior.
In fact, the role of income in driving recycling participation is
less clear than expected. The income effects could run in different
directions (Miafodzyeva and Brandt, 2013). Hage and So € derholm
(2008), for example, argue that wealthier households may lose
motivation to the extent in which their opportunity cost of recy-
cling participation (arising from sorting, storage, and transferring,
etc.) is higher than lower-income groups. This view has found
empirical support from Sidique et al. (2010) report, which was
based on actual recycling rates recorded in Minnesota. Chung and
Poon (2001) offer another explanation in a developing-country
context: lower income groups are keen on recycling because of
the economic returns from selling off high-value recyclables. Their
opportunity cost of recycling participation tends to be lower, rela-
tive to other everyday priorities.
Despite their deep insights, most of these international studies
fail to account for the spatial constraints of compact cities that are
filled with an incredible mix of high-rise, high-density de-
velopments, such as Hong Kong. In these cities, curbside collection
of recyclables is familiar and available to only a minority of
householders. The great majority drop waste items in designated
recycling facilities at a central location and adapt to a different
combination of spatial conditions in which these facilities are sit-
uated, which are often shaped by estate management. Moreover,
few studies have directly addressed problems associated with using
the self-reporting approach. An exception is Yau (2010), who in-
vestigates how actual recycling outcomes are related to income and
other socio-economic characteristics using a sample of 122 private
residential estates in Hong Kong. He shows that income and age are
key predictors of recycling behavior in a high-rise, high-density
residential setting.
The present research attempts to address an apparent limitation
of Yau (2010) approach, i.e. the failure to represent public rental
housing estates. The importance of socio-economic profile in
determining recycling outcomes brings to the forefront the policy
bias mentioned in the last section. Lower-income communities
should be given more policy and financial support if recycling
outcomes improve with higher household incomes and more
generally socio-economic status. Analyzing a representative sample
420 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425
of public rental housing estates can help identify areas requiring
preferential arrangements.
4. Methods
4.1. Data
We gathered two sets of official records for this study, one from
the EPD and the other from CSD. A regression analysis was con-
ducted, using dependent variables derived from records of do-
mestic waste recycling. The EPD received these records directly
from the housing estates or buildings participating in the SSWP
Programme, which have been published on the SSWP website since
May 2017.2 Waste recycling companies collected recyclables from
each participating housing estate on a regular basis. They recorded
the volume of recyclables collected on-site and issued receipts to
the residential property management agent. When submitting
these records to the EPD, property managers wrote to confirm that
all recorded recyclables been transferred to a recycling company for
recycling purposes, and that the information provided be true,
correct and complete. Furthermore, the original receipts issued by
the third party, i.e. recycling companies, were subject to verification
by the EPD. These records are therefore considered a reliable source
of data for the present research.
The recycling data used for this study cover a twelve-month
period in 2016. A total of 293 housing estates across the city of
Hong Kong offered consent to the disclosure of the domestic
recycling records they submitted to the EPD for this period.3 Each
record published by the EPD represents recycling intensity, defined
as the monthly average volume of recyclables (mainly paper,
plastic, and metal) collected from a residential estate during the
twelve-month period (Yau, 2010). Most of the 293 estates are large
residential complexes located in the city's built-up areas, including
187 public rental housing estates established by the Hong Kong
Housing Authority (HA) or the Hong Kong Housing Society (HS).
The rest of them are predominantly private residential premises,
along with a small number of student dormitories and publicly
subsidized housing estates designated for private ownership under
the Government's Home Ownership Scheme. As our study specif-
ically focused on publicly owned and rental residential estates, only
these 186 observations were enlisted for this inquiry.
As Yau (2010) has indicated, the absolute volume of recyclables
is not a perfect indicator, because it does not account for the total
volume of domestic waste generated. The amount of the former
could be inflated by an increase in the latter, which does not involve
waste minimization and may result in a larger amount of waste
being transferred to the overloaded landfills in Hong Kong. We did
not find any information on the amount of domestic waste gener-
ated or disposed in each housing estate. The closest we could obtain
was the amount of domestic waste disposed in each Council District
(Environmental Protection Department, 2016). Although these
district-level estimates are far from ideal, due to a large number of
housing estates in some Districts complicating the interpretation of
regional averages, they provide a close approximation of a
contextual factor, i.e. total volume of waste generated. An alterna-
tive dependent variable was accordingly constructed by dividing
waste recovery by waste disposal, called ‘relative recycling intensity’.
2
A full list of participating housing estates or buildings is available from www.
wastereduction.gov.hk/en/household/source_achievements.htm. Accessed 19 June
2017.
3 where Intensity ðRÞi indicates the volume of recyclables collected
A full list of the 293 housing estates or buildings is available from www.
wastereduction.gov.hk/sites/default/files/en/materials/household/Result%20list_
2015-16_Eng_p.xls. Accessed 19 June 2017.
Since generation quantity (g) is the sum of recovery quantity (r) and
disposal quantity (d), relative recycling intensity equals grr and
changes with the amount of domestic waste not recovered. Despite
its methodological imperfections, this measure takes into consid-
eration the fact that the amount of recyclables tends to increase
with waste generation.
Our analysis followed the approach adopted by Yau (2010),
which is deemed to be appropriate and straightforward, consid-
ering the sample size and the properties of variables. The depen-
dent variables were regressed on a suite of socio-economic and
demographic characteristics of local population. Information about
these characteristics was solicited from the latest full population
census administered by the CSD. The census provides population
statistics at four planning scales, namely, District Council Districts,
Constituency Areas, Tertiary Planning Units, and Major Housing
Estates. The CSD defines a Major Housing Estate as a group of res-
idential buildings developed by the same developer (either in the
public sector or in the private sector) in one or more phases in a
neighborhood and with at least 3000 residents or 1000 domestic
households. We retrieved population statistics at this planning
scale because the volume of recyclables collected was presented at
the estate/building level. Twenty-nine SSWP participating estates
were established after the census period or too small to be classi-
fied as Major Housing Estates. These 29 cases were therefore
removed from our sample, leaving 158 cases for statistical analysis.
Key population characteristics of individual Major Housing Es-
tates were extracted for this analysis, relating to population,
gender, age, household income, household size, and the proportion
of working population in housing estate. All of them were derived
from the census database. The type of property management agent
(i.e. private company or public agency) was also included in anal-
ysis. This information was associated with the domestic waste
recycling records published by the EPD. A dummy variable is
created, which takes 1 if a housing estate was managed by a private
property management company, or 0 if otherwise case. These
variables could ascertain what effects socio-economic and de-
mographic factors might have on waste recycling intensity. In
addition, we controlled for location effects by including dummy
variables representing the district in which a housing estate is
located.
4.2. Estimation
Model estimation was predominantly based on continuous
variables. The Ordinary Least Squares (OLS) regression was used in
the statistical analysis. The regression model for absolute recycling
intensity is specified as:


ln Intensity ðAÞi ¼ a þ bj xji þ … þ bp xpi þ bd di þ ε1i (1)
where Intensity ðAÞi indicates the volume of recyclables collected
from housing estate i. a is the intercept. xji… xpi denotes socio-
economic and demographic characteristics of the housing estate
j…p, including, population, median age, median household income
and so forth. di denotes the district in which the housing estate is
located. ε1i is the error term.
The regression model for relative recycling intensity is specified
as:
Intensity ðRÞi ¼ a þ bj xji þ … þ bp xpi þ bd di þ ε2i (2)
from housing estate i, relative to the volume of domestic waste
collected from its home district for disposal. All other variables are
 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425
Fig. 3. Locations of the public rental housing estates in the sample.
as defined above.
5. Results
5.1. Descriptive statistics
The sample includes 158 public rental Major Housing Estates in
Hong Kong. Fig. 3 indicates their locations. As shown in Table 2,
they are concentrated in densely built, mixed land use areas, such
as Kwun Tong and Kwai Tsing Districts, where the two largest
housing estates in the sample, with more than 36,000 residents
each, are located. At the bottom of the table are Tai Po and Central
and Western Districts. While Central and Western is essentially a
central business district with a handful of public residential units,
the three housing estates in Tai Po have a population of more than
15,000 each, exceeding the average. The sampled estates are home
to a relatively large cohort of dwellers, with a population size of 1.9
million in 2011. A total of 0.66 million domestic households
inhabited these Major Housing Estates (Census and Statistics
Department, 2012), representing 27.8 per cent of all domestic
households in Hong Kong (2.37 million) and 91.5 per cent of those
registered in the public rental housing system (0.72 million). On
average, each of these Major Housing Estates accommodated 4173
households. Median internal floor area ranged from 16.3 m2
(smallest) to 52.2 m2 (largest) (Hong Kong Housing Authority,
2017).
Other socio-economic characteristics of these Major Housing
Estates are displayed in Table 3, which includes all of the model
variables. Each estate had an average of approximately twelve
thousands residents (12,285), each living in a family of three (2.9).
Seventeen per cent of these residents were aged 65 or above, while
421
the median age was 43.8. Over half of them were females (52 per
cent). Slightly more than one-tenth (11.3 per cent) of those aged 20
or above and not being students had attained post-secondary ed-
ucation. Less than half (43.5 per cent) were either employees or
employers, putting them in the working population. These two
categories, as defined by the CSD, are “employees/unpaid family
workers” and “employers/self-employed” respectively. Both are
included as “working population”. Non-working population in-
cludes home-makers, students, and retirees. The median renting
household made HK$13,100 per month (approx. US$1700). More
than half of these residential sites (53.8 per cent) were managed by
private property management companies, whereas others were
overseen by public agencies (i.e. the HA or HS). The population and
income variables were log transformed in the following regression
analysis.
The volume of recyclables collected from these 158 Major
Housing Estates ranged from 0.19 kg to 20.24 kg per month per
household, with a mean value of 3.9 kg and a median of 3.0 kg
(Table 3). The mean value increased by a small margin to 4.03 kg
when the other 29 SSWP participating public housing estates were
included in calculation. It is much lower than that of the 83 private
residential estates in our full sample, which averaged 6.59 kg,4 and
Yau (2010, p. 2445) estimate of 5.52 kg (or 66.22 kg in a full year)
across 122 private residential estates in 2008. The relative recycling
4
There were 293 cases in the list of SSDW participating housing estates with
recycling statistics. Five of these observations were submitted by different phases of
two housing estates. Those from the same estates were merged into one group
(normalized for number of units), reducing the five cases to two. As a result, the
adjusted full sample has 290 cases, including 187 public rentals, 83 private estates,
15 estates under the Home Ownership Scheme (i.e. not being public rentals), 4
student dormitories, and 1 mixed ownership mode (public rental and flat-for-sale).
422 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425

Table 2
Geographical distribution of sample.

District Case Average population Average number of households

Frequency Percentage (%)

Kwun Tong 28 17.7 11,964 4220

Kwai Tsing 19 12 14,760 4968
Wong Tai Sin 15 9.5 12,832 4319
Yuen Long 14 8.9 12,869 4057
Eastern 13 8.2 8869 2967
Sha Tin 13 8.2 12,038 4077
Sham Shui Po 11 7 10,923 3902
Tuen Mun 9 5.7 14,494 5491
Southern 7 4.4 12,934 4268
Tsuen Wan 7 4.4 9271 3233
Kowloon City 6 3.8 9595 3321
North 5 3.2 11,883 4062
Sai Kung 4 2.5 14,694 4494
Islands 3 1.9 15,053 4277
Tai Po 3 1.9 16,662 5408
Central and Western 1 0.6 4199 1547
All 158 100 12,285 4173

Notes: all figures refer to the selected public rental housing estates, and not the entire territory nor Districts. Two Districts (i.e. Yau Tsim Mong and Wan Chai) are not
represented in the sample.

Table 3
Descriptive statistics.

Variable Description Mean Median Min Max S.D. N

POPULATION Total population of housing estate (persons) 12,285 11,699 2914 37,547 6342 158
AGE Median age of residents (years) 43.8 44.7 28.2 54.0 5.4 158
FEMALE Female residents (%) 52.0 51.8 46.9 56.7 1.4 158
EDUCATION Proportion of residents aged 20 or above and not being students had attained post-secondary education (%) 11.3 10.7 4.8 29.2 3.0 158
HHSIZE Number of household members (persons) 2.9 2.9 1.5 4.3 0.4 158
WORKPOP Working population (%) 43.5 43.1 34.7 58.8 3.8 158
INCOME Median monthly income (all domestic households) (‘000 HK$) 13.1 13.0 6.5 20.3 2.8 158
PRIVATE Property management (1 if a private company, 0 if otherwise) 0.5 1.0 0.0 1.0 0.5 158
INTENSITY(A) Absolute volume of recyclables collected from housing estate (kg per month per household) 3.9 3.0 0.1 20.2 3.5 158
INTENSITY(R) Volume of recyclables collected divided by volume of domestic waste disposed at District 1.6 1.2 0.04 9.3 1.3 158

intensity, defined as the ratio of waste recovery to waste disposal,
ranged from 0.04 to 9.28. It has a mean value of 1.6 and a median of
1.2, suggesting that for every one unit of disposed waste, 1.6 units of
waste was recycled and recovered. This may be an over-estimate,
probably due to inherent selection bias, i.e. SSWP participating
housing estates were better supported and more active in pro-
moting waste recycling, and low achieving estates did not submit
their recycling statistics.
5.2. Regression analysis
Table 4 presents the correlation matrix for all variables. It shows
that the log-transformed absolute recycling intensity (i.e. volume of
recyclables collected from public rental housing estates) was a
function of residents' median age, household size, median house-
hold income, and private property management, when other fac-
tors were not controlled for. As a derivative of absolute quantity,
relative recycling intensity also correlated with these variables in
the same way. Population, gender, education, proportion of work-
ing population had limited effects on the intensity variables.
Table 5 shows two statistical models regressing absolute recy-
cling intensity on the socio-economic and demographic variables
listed in Table 3. Both models controlled for location (district) ef-
fects. The Durbin-Watson statistics fall between 1.5 and 2.5, ruling
out the possibility of autocorrelation. The first model has an
adjusted R2 of 0.17, suggesting that it explains 17 per cent of vari-
ance in the dependent variable. PRIVATE creates positive impacts
on absolute recycling intensity. Public rental housing estates
managed by a private property management company are more
productive in waste recycling and separation than those managed
by a public agency. Other independent variables do not achieve
statistical significance. These include median age, which was found
to be a key predictor in a similar study in Hong Kong (Yau, 2010).
The second model is created by removing seven cases from the
previous one. These seven cases yield a Cook's distance greater than
the cut-off value (4/N) and therefore deemed to be problematic.
Cook's distance identifies cases that are influential or have a large
effect on the regression solution and may be distorting the solution
for the remaining cases in the analysis. The higher the Cook's dis-
tance is, the more influential a particular observation is. The con-
ventional cut-off point is 4/Nein this case, 4/158. With a reduced
number of observations, this model returns an improved adjusted
R2 (0.284). The positive effect of ln (INCOME) becomes significant.
This indicates that the volume of recyclables collected from public
rental housing estates increases with household income and the
presence of private management.
The two models displayed in Table 6 include relative recycling
intensity as the dependent variable, which accounts for the likely
association between waste recovery and disposal. Model 3, with the
complete sample, explains 19 per cent of variance in relative
recycling intensity, and PRIVATE remains the only significant factor.
Based on Cook's distance, seven outliners (different from the pre-
vious ones) are removed from the analysis. Model 4, with 151 ob-
servations, has an adjusted R2 of 0.231, indicating an improvement
in explanatory power. PRIVATE and ln (INCOME) are positively
associated with the dependent variable. This means that the
 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425 423

Table 4
Correlation matrix for all variables.

Variable Correlation coefficient

2 3 4 5 6 7 8 9 10

1. ln (INTENSITY(A)) .819** 0.109 .266** 0.072 0.137 .337** 0.129 .315** .311**
2. INTENSITY(R) e 0.062 .171* 0.03 0.144 .304** 0.096 .287** .333**
3. ln (POPULATION) e 0.148 .184* .189* 0.063 0.003 0.015 0.03
4. AGE e .360** 0.106 .655** 0.064 .250** .347**
5. FEMALE e 0.021 .282** .292** .226** 0.053
6. EDUCATION e .240** .446** .495** 0.126
7. HHSIZE e .322** .779** .240**
8. WORKPOP e .650** 0.039
9. ln (INCOME) e 0.103
10. PRIVATE e

Pearson correlation (two-tailed). *p < .05, **p < .01.

Table 5
OLS regression on the volume of recyclables collected from 158 Major Housing Estates.

Model 1 Model 2

Coefficient Standard Error 95% Confidence Coefficient Standard Error 95% Confidence Interval
Interval

(Constant) 0.814 4.345 7.779 9.408 0.633 3.722 7.998 6.732

ln (POPULATION) 0.182 0.149 0.476 0.112 0.001 0.13 0.256 0.257
AGE 0.045 0.031 0.105 0.016 0.032 0.026 0.082 0.019
FEMALE 0.029 0.068 0.106 0.164 0.022 0.058 0.093 0.137
EDUCATION 0.017 0.033 0.048 0.083 0.005 0.038 0.081 0.071
HHSIZE 0.479 0.568 1.603 0.645 0.498 0.473 1.434 0.437
WORKPOP 0.033 0.036 0.104 0.039 0.032 0.031 0.093 0.029
ln (INCOME) 1.789 0.996 0.18 3.758 1.841* 0.848 0.163 3.519
PRIVATE 0.451* 0.178 0.098 0.804 0.661** 0.150 0.364 0.958
(Control for District) Yes Yes
Adj. R2 0.170 0.284
Durbin-Watson 1.901 1.964
F 2.400 3.581
Sig. 0.001 0.000
Number of obs. 158 151

*p < .05. **p < .01. Dependent variable: ln (INTENSITY(A)).

Table 6
OLS regression on relative recycling intensity.

Model 3 Model 4

Coefficient Standard Error 95% Confidence Coefficient Standard Error 95% Confidence Interval
Interval

(Constant) 5.567 5.506 16.457 5.324 4.479 4.478 4.383 13.341

ln (POPULATION) 0.010 0.189 0.363 0.383 0.299* 0.146 0.589 0.010
AGE 0.016 0.039 0.093 0.061 0.023 0.031 0.084 0.038
FEMALE 0.085 0.086 0.086 0.255 0.018 0.069 0.154 0.118
EDUCATION 0.037 0.042 0.046 0.120 0.022 0.044 0.108 0.064
HHSIZE 0.139 0.720 1.564 1.286 0.294 0.569 1.420 0.833
WORKPOP 0.045 0.046 0.136 0.046 0.061 0.037 0.135 0.013
ln (INCOME) 1.881 1.262 0.614 4.377 2.062* 0.995 0.094 4.031
PRIVATE 0.869** 0.226 0.421 1.316 0.517** 0.176 0.168 0.866
(Control for District) Yes Yes
Adj. R2 0.190 0.231
Durbin-Watson 1.862 1.867
F 2.603 2.963
Sig. 0.000 0.000
Number of obs. 158 151

*p < .05. **p < .01. Dependent variable: Intensity(R).

volume of recyclables collected, relative to disposed domestic 6. Discussion

waste, increases with household income and the presence of pri-
vate management. ln (POPULATION) is also found to be statistically
significant. Yet, care is required when interpreting the population
effect as this variable achieves significance only in the last model
and does not correlate with relative recycling intensity when other
variables are not controlled for.
The findings presented above provide an alternative empirical
basis for re-visiting existing knowledge about household recycling
behavior. Most of the international and local studies about waste
recycling behavior are based on self-reports and individual-level
analysis, predominantly in a low- or medium-density milieu. The
424 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425

main contribution of the present study lies in its community-based inevitably compromised prediction. Availability of waste disposal
and outcome-based approach, using the entire housing estate as a estimates at a sub-district level will substantially improve the
unit of analysis and analyzing actual recycling outcomes. Only a quality of future research.
handful of researchers, such as Sidique et al. (2010) and Yau (2010), The positive income effect has implications for policy practice.
have adopted such an approach. Yet, their pioneering attempts are Using official statistics, Ko and Poon (2009, p. 106) have shown that
subject to sampling constraints (e.g. being restricted to low/ domestic waste disposal in Hong Kong increased with GDP before
medium-density or private residential areas). The present study 1998, but the correlation between them deteriorated after 1998.
contributes to methodological development by demonstrating that They attribute the poorer correlation post 1998 to the introduction
this approach is applicable to a different context, giving impetus to of waste recovery policies. While these policies achieved partial
wider applications in similar, highly urbanized regions. success territory-wide, variations in recycling outcomes across
The findings have indicated variations from other studies. For residential communities continue to exist. The positive income
example, while many individual-level analyses on self-reported effect might suggest that economically disadvantaged residential
behavior suggest that recyclers tend to be mature and educated communities encounter greater difficulties in promoting waste
(Belton et al., 1994; Tucker et al., 1998; Vencatasawmy et al., 2000; separation and recycling. On the other hand, their better off private
Perrin and Barton, 2001; Martin et al., 2006), there is little evidence counterparts appear to be more active, a looming indication of the
from the present research. The role of gender (Chan, 1998; Chung middle-class environmentalism prevalent in Hong Kong back in
and Poon, 2001; Ekere et al., 2009) remains unclear. The socio- 1990s (Chan and Hills, 1993; Lo, 2016a).
economic effects on recycling outcomes are limited, with the
exception of household income. Previous studies conducted by
Sidique et al. (2010) and Yau (2010) have confirmed the importance
7. Conclusions
of income. Unlike Sidique et al. (2010), however, Yau (2010) and the
present research (both reporting on Hong Kong) indicate a mod-
Like many other high-density cities, Hong Kong has been
erate positive income effect. Considering that Sidique et al. (2010)
struggling to find space for disposing municipal solid waste. Local
collected their data from a sample of US counties, the contrasting
waste recovery initiatives rarely account for the variations in resi-
results might be an indication for the role of urban form in medi-
dents' socio-economic profiles across place, leading to mixed out-
ating the income-recycling relationship: does the compactness of
comes among diverse communities. This study investigated socio-
residential communities affects the direction of income effect on
economic effects on recycling behavior in a sample of subsidized
waste recycling behavior? Additional information, such as urban or
rental housing estates. We found that recycling outcomes vary with
neighborhood density, location of recycling facilities and their
a limited set of socio-economic factors. Housing estates managed
distance from homes, would be useful for exploring the contested
by a private property management company and populated by
role of income.
better off households collected more recyclables from their resi-
Although the study focused on public rentals only and did not
dents. Measures of absolute and relative recycling intensity ach-
explore the role of dwelling type (Martin et al., 2006; Miafodzyeva
ieved similar results.
and Brandt, 2013), the average recycling amount suggests that
The positive income effect warrants policymakers' attention.
public housing estates fall behind the private ones. This provides an
Introducing demographically differentiated waste recovery policies
initial evidence on the effect of housing type, which is broadly
could mitigate such socio-economic tendencies. Our policy rec-
consistent with a common observation among international
ommendations include, for example, distributing more site-specific
studies that homeowners are more likely to be active recyclers than
recycling bins in economically disadvantaged residential commu-
renters (Vencatasawmy et al., 2000; Perrin and Barton, 2001). A
nities, and reversing the ECF's funding rule that precludes public
closely related finding is the strong correlation between recycling
rental housing estates from submitting a funding application.
quantity and the type of property management agent. Private
Additional concessions towards management fees for these estates,
companies have an incentive to organize a variety of participatory
particularly those managed by a public agency, might also be
activities within their remit, such as waste separation and recy-
considered. This could create more economic incentives for resi-
cling, because their performance as perceived by residents and
dents and estate management to engage in waste recycling and
their representatives is often linked to outstanding management
separation.
outcomes, which might include a recycling award from a trusted
A limitation of this study is that there is no information on the
institution, e.g. EPD or ECF. Public agencies have lower motivation
availability of waste recycling and separation facilities in the
to act. More efforts are therefore needed to promote waste recy-
housing estates, which might affect residents' participation and
cling in public rental housing estates, particularly those currently
recycling rates. Another unknown is the residents' attitude towards
managed by a public agency.
waste recycling and separation. Both sets of information could be
The present study has demonstrated a positive income effect.
solicited from a series of questionnaire surveys, which are therefore
There is a significant association relationship between household
recommended for future research along with an analysis of actual
income and recycling intensity, corroborating a number of previous
recycling outcomes.
studies suggesting that wealthier families and individuals are more
likely to participate in domestic waste recycling (Gamba and
Oskamp, 1994; Mak et al., 2017; Martin et al., 2006; Owens et al.,
2000; Perrin and Barton, 2001; Vencatasawmy et al., 2000). It Acknowledgement
shows consistency with the post-materialist tendency, implying
that higher recycling intensity might be driven by changes in The authors thank the editors and the four anonymous re-
people's values and attitudes towards environmentally responsible viewers of the Journal of Environmental Management for their
directions. Further evidence on values and attitudes is needed to useful comments and suggestions. The generosity of the Environ-
validate this claim. Another possibility is the parallel increase in mental Protection Department for making the useful data openly
general consumption (hence, waste generation) with household available is highly appreciated. The first author received funding
income. We have accounted for the amount of domestic waste support for this study from the Environment and Conservation
disposal, but the lack of access to estate-level disposal records has Fund (Project Reference Number: TF02/2015).
 A.Y. Lo, S. Liu / Journal of Environmental Management 214 (2018) 416e425 425

References Gelissen, J., 2007. Explaining popular support for environmental protection: a
multilevel analysis of 50 nations. Environ. Behav. 39, 392e415.
Hage, O., So €derholm, P., 2008. An econometric analysis of regional differences in
Belton, V., Crowe, D.V., Matthews, R., Scott, S., 1994. A survey of public attitudes to
household waste collection: the case of plastic packaging waste in Sweden.
recycling in Glasgow (U.K.). Waste Manag. Res. 12, 351e367.
Waste Manag. 28, 1720e1731.
Brechin, S.R., 1999. Objective problems, subjective values, and global environmen-
Hong Kong Housing Authority, 2017. Property Location and Profile. Hong Kong
talism: evaluating the postmaterialist argument and challenging a new expla-
Housing Authority, Hong Kong. Available from. www.housingauthority.gov.hk/
nation. Soc. Sci. Q. 80, 793e809.
en/global-elements/estate-locator/index.html. (Accessed 19 June 2017).
Census and Statistics Department, 2002. Thematic Household Survey Report No. 9.
Inglehart, R., 1990. Culture Shift in Advanced Industrial Society. Princeton Univer-
Hong Kong SAR Government, Hong Kong.
sity Press, Princeton,NJ.
Census and Statistics Department, 2012. Population Census. Hong Kong SAR Gov-
Kim, S.Y., Wolinsky-Nahmias, Y., 2014. Cross-national public opinion on climate
ernment, Hong Kong, p. 2011.
change: the effects of affluence and vulnerability. Glob. Environ. Polit. 14,
Chan, C., Hills, P.R., 1993. Limited Gains: Grassroots Mobilization and the Environ-
79e106.
ment in Hong Kong Centre of Urban Planning and Environmental Management.
Ko, P.S., Poon, C.S., 2009. Domestic waste management and recovery in Hong Kong.
University of Hong Kong, Hong Kong.
J. Mater. Cycles Waste Manag. 11, 104e109.
Chan, K., 1998. Mass communication and pro-environmental behaviour: waste
Lee, S., Paik, H.S., 2011. Korean household waste management and recycling
recycling in Hong Kong. J. Environ. Manag. 52, 317e325.
behavior. Build. Environ. 46, 1159e1166.
Cheung, L.T.O., Chow, A.S.Y., Fok, L., Yu, K.-M., Chou, K.-L., 2017. The effect of self-
Lima, M.L., Barnett, J., Vala, J., 2005. Risk perception and technological development
determined motivation on household energy consumption behaviour in a
at a societal level. Risk Anal. 25, 1229e1239.
metropolitan area in southern China. Energy Effic. 10, 549e561.
Lo, A.Y., 2014. Negative income effect on perception of long-term environmental
Chung, S.S., Leung, M.M.Y., 2007. The value-action gap in waste recycling: the case
risk. Ecol. Econ. 107, 51e58.
of undergraduates in Hong Kong. Environ. Manag. 40, 603e612.
Lo, A.Y., 2016a. Public discourses of climate change in Hong Kong. J. Environ. Policy
Chung, S.S., Poon, C.S., 1994. Hong Kong citizens' attitude towards waste recycling
Plan. 18, 27e46.
and waste minimization measures. Resources, Conserv. Recycl. 10, 377e400.
Lo, A.Y., 2016b. Small is green? Urban form and sustainable consumption in selected
Chung, S.S., Poon, C.S., 2001. A comparison of waste-reduction practices and new
OECD metropolitan areas. Land Use Policy 54, 212e220.
environmental paradigm of rural and urban Chinese citizens. J. Environ. Manag.
Mak, K.B., Cheung, T.L., Hui, L.D., 2017. Community participation in the decision-
62 (1), 3e19.
making process for sustainable tourism development in rural areas of Hong
Coggins, P., 1994. Who is the recycler? J. Waste Manag. Resour. Recovery 1, 69e75.
~es, P., Marques, R.C., 2014. Packaging waste Kong, China. Sustainability 9.
da Cruz, N.F., Ferreira, S., Cabral, M., Simo
Martin, M., Williams, I.D., Clark, M., 2006. Social, cultural and structural influences
recycling in Europe: is the industry paying for it? Waste Manag. 34, 298e308.
on household waste recycling: a case study. Resour. Conserv. Recycl. 48,
Dunlap, R.E., York, R., 2008. The globalization of environmental concern and the
357e395.
limits of the postmaterialist values explanation: evidence from four multina-
Miafodzyeva, S., Brandt, N., 2013. Recycling behaviour among householders: syn-
tional surveys. Sociol. Q. 49, 529e563.
thesizing determinants via a meta-analysis. Waste Biomass Valorization 4,
Ekere, W., Mugisha, J., Drake, L., 2009. Factors influencing waste separation and
221e235.
utilization among households in the Lake Victoria crescent, Uganda. Waste
Miliute-Plepiene, J., Hage, O., Plepys, A., Reipas, A., 2016. What motivates house-
Manag. 29, 3047e3051.
holds recycling behaviour in recycling schemes of different maturity? Lessons
Environment and Conservation Fund, 2016. Guide to Application - Programme on
from Lithuania and Sweden. Resour. Conserv. Recycl. 113, 40e52.
Source Separation of Domestic Waste Environment and Conservation Fund,
Owens, J., Dickerson, S., Macintosh, D.L., 2000. Demographic covariates of resi-
Hong Kong.
dential recycling efficiency. Environ. Behav. 32, 637e650.
Environment Bureau, 2013. Hong Kong Blueprint for Sustainable Use of Resources
Pakpour, A.H., Zeidi, I.M., Emamjomeh, M.M., Asefzadeh, S., Pearson, H., 2014.
2013-2022. Hong Kong SAR Government, Hong Kong.
Household waste behaviours among a community sample in Iran: an applica-
Environmental Protection Department, 1997. Environment Hong Kong 1997. Help
tion of the theory of planned behaviour. Waste Manag. 34, 980e986.
Us Make a Better World. Hong Kong SAR Government, Hong Kong.
Perrin, D., Barton, J., 2001. Issues associated with transforming household attitudes
Environmental Protection Department, 2005. A Policy Framework for the Man-
and opinions into materials recovery: a review of two kerbside recycling
agement of Municipal Solid Waste (2005-2014). Hong Kong SAR Government,
schemes. Resour. Conserv. Recycl. 33, 61e74.
Hong Kong.
Planning Department, 2015. Planning Data: Land Utilization in Hong Kong 2015.
Environmental Protection Department, 2015. Monitoring of Solid Waste in Hong
Hong Kong Government, Hong Kong. Available at. http://www.pland.gov.hk/
Kong: Waste Statistics for 2014. Hong Kong SAR Government, Hong Kong.
pland_en/info_serv/statistic/landu.html. (Accessed 19 June 2017).
Environmental Protection Department, 2016. Monitoring of Solid WasteIn Hong
Sidique, S.F., Joshi, S.V., Lupi, F., 2010. Factors influencing the rate of recycling: an
Kong: Waste Statistics for 2015. Hong Kong SAR Government, Hong Kong.
analysis of Minnesota counties. Resour. Conserv. Recycl. 54, 242e249.
Environmental Protection Department, 2017. Monitoring of Solid Waste in Hong
Simo~ es, P., Marques, R.C., 2012. On the economic performance of the waste sector. A
Kong: Waste Statistics for 2016. Hong Kong SAR Government, Hong Kong.
literature review. J. Environ. Manag. 106, 40e47.
Franzen, A., Meyer, R., 2010. Environmental attitudes in cross-national perspective:
Thi, N.B.D., Kumar, G., Lin, C.-Y., 2015. An overview of food waste management in
a multilevel analysis of the ISSP 1993 and 2000. Eur. Sociol. Rev. 26, 219e234.
developing countries: current status and future perspective. J. Environ. Manag.
Franzen, A., Vogl, D., 2013. Two decades of measuring environmental attitudes: a
157, 220e229.
comparative analysis of 33 countries. Glob. Environ. Change 23, 1001e1008.
Tucker, P., Murney, G., Lamont, J., 1998. Predicting recycling scheme performance: a
Gamba, R.J., Oskamp, S., 1994. Factors influencing community residents' participa-
process simulation approach. J. Environ. Manag. 53, 31e48.
tion in commingled curbside recycling programs. Environ. Behav. 26, 587e612. €
Vencatasawmy, C.P., Ohman, M., Bra€nnstro€ m, T., 2000. A survey of recycling
Gamberini, R., Del Buono, D., Lolli, F., Rimini, B., 2013. Municipal solid waste
behaviour in households in Kiruna, Sweden. Waste Manag. Res. 18, 545e556.
management: identification and analysis of engineering indexes representing
Yau, Y., 2010. Domestic waste recycling, collective action and economic incentive:
demand and costs generated in virtuous Italian communities. Waste Manag. 33,
the case in Hong Kong. Waste Manag. 30, 2440e2447.
2532e2540.