Professional Documents
Culture Documents
DOI 10.1007/s10163-008-0233-1
Springer 2009
Introduction
nomic growth in the past few decades by shifting the economic structure from the agriculture sector to the industrial
and service sectors. Additionally, there was liberalization in
the social and political environment after the abrogation of
martial law in 1987. Industrialization and urbanization diffused rapidly throughout Taiwan, causing dramatic changes
in lifestyle and consumer behavior.
Some socioeconomic indices in terms of lifestyle changes
are examined. The official database1 was used for analysis
of the period from 1984 to 2005, and the indices analyzed
in this study are described in Table 1. Figure 1 demonstrates
the normalized trends of these indices, and the high correlation coefficients (see Table 2) among the indices confirm
their intrinsic interaction, which may result from the similarities in overall lifestyle. Such evolution may be regarded
as a characteristic of a developing country. Figure 2 shows
the time series data of per capita consumption expenditure
by category from 1984 to 2005; all the consumption expenditures changed significantly with changes in lifestyle.
However, as recent studies have indicated, this rapidly
increasing production and consumption may not be happening in a sustainable way and is likely to lead to global
environmental problems.25 The consumer behavior originating from the contemporary lifestyle has a great potential
for increasing environmental loads, particularly municipal
solid waste (MSW) generation and discard.69 In this study,
MSW discards is defined as the amount of MSW generation
that will be regularly collected and treated by the local
municipalities.
Abstract Industrialization and urbanization result in significant changes in lifestyle. These lifestyle changes seem to
lead to unsustainable consumption patterns and increase
the generation of various kinds of environmental loads,
especially the amount of municipal solid waste (MSW).
Taiwan is a small island with scarce natural resources. The
economic development in Taiwan has resulted in the generation of large amounts of MSW. As a result, the Taiwan
Environmental Pollution Administration (TEPA) has produced regulations for waste minimization and has imposed
several important policy measures that have successfully
reduced the MSW discard rate in recent years and have
established a public recycling network as a part of the MSW
collection. Nowadays, the objective of the MSW policies in
Taiwan is to develop a zero-waste society. This article
aims to review the MSW management progress in Taiwan
and to project future MSW discards up to 2011 based on
the national plan and assumed scenarios for socioeconomic
variables. According to the analysis results, a more sustainable consumption pattern can be proposed and the corresponding MSW management system can be planned so as
to develop a low-waste-discard society.
Key words Municipal solid waste management Developing country Changes in lifestyle Consumer behavior
Econometric modeling
As the socioeconomic environment changes, the MSW generation rate also changes to a great extent. In this study,
MSW is classified into eight categories by physical property
for treatment and final disposal planning. Since the MSW
composition analysis was conducted on a dry basis before
2004, the moisture of waste is separated and taken as a
single category (however, the analysis changed to a wet
111
Table 1. Description of the representative indices related to changes in lifestyle in Taiwan
Index
Description
Unit
Poput
Hhldt
Hlbrrt
Hpopt
Aveaget
Hun15rt
Hov65rt
SFt
WKt
Ginit
Engelt
Unempt
Savingt
The
The
The
The
The
The
The
The
The
The
The
The
The
Capita
Household
%
Capita
Years
%
%
%
%
%
%
%
%
Hhldt
Hpopt
Hlbrrt
Aveaget
Hun15rt
Hov65rt
SFt
WKt
Ginit
Engelt
Unempt
Savingt
Hhldt
Hpopt
Hlbrrt
Aveaget
Hun15rt
Hov65rt
SFt
WKt
Ginit
Engelt
Unempt
Savingt
1.000*
0.995*
0.995*
0.975*
0.993*
0.998*
0.997*
0.985*
0.943*
0.955*
0.962*
0.664*
0.995*
1.000*
0.994*
0.971*
0.997*
0.998*
0.996*
0.974*
0.962*
0.959*
0.942*
0.722*
0.995*
0.994*
1.000*
0.980*
0.988*
0.994*
0.997*
0.988*
0.938*
0.957*
0.967*
0.661*
0.975*
0.971*
0.980*
1.000*
0.971*
0.977*
0.979*
0.979*
0.944*
0.931*
0.951*
0.616*
0.993
0.997
0.988
0.971
1.000
0.996
0.994
0.971
0.967
0.958
0.934
0.723*
0.998*
0.998*
0.994*
0.977*
0.996*
1.000*
0.998*
0.978*
0.955*
0.949*
0.953*
0.692*
0.997*
0.996*
0.997*
0.979*
0.994*
0.998
1.000*
0.980*
0.946*
0.953*
0.964*
0.681*
0.985*
0.974*
0.988*
0.979*
0.971*
0.978*
0.980*
1.000*
0.909*
0.940*
0.969*
0.567*
0.943*
0.962*
0.938*
0.944*
0.967*
0.955*
0.946*
0.909*
1.000*
0.931*
0.839*
0.813*
0.955*
0.959*
0.957*
0.931*
0.958*
0.949*
0.953*
0.940*
0.931*
1.000*
0.907*
0.736*
0.962*
0.942*
0.967*
0.951*
0.934*
0.953*
0.964*
0.969*
0.839*
0.907*
1.000*
0.498
0.664*
0.722*
0.661*
0.616*
0.723*
0.692*
0.681*
0.567*
0.813*
0.736*
0.498
1.000
3
Poput
Aveaget
Hhldt
Hlbrrt
1
Hpopt
Hun15rt
Hov65rt
SFt
WKt
-1
Ginit
Engelt
Unempt
-2
Savingt
-3
1984
1987
1990
basis in 2005). Based on such classification, essential information can be obtained for designing recycling, treatment,
and disposal plans.
Figure 3 illustrates the official records of the composition
of discarded MSW in Taiwan. The dataset covers the period
19922004, whereas per capita overall MSW discard levels,
which denote the overall amount of discard for all the waste
1993
1996
1999
2002
2005
fractions, are available from 1988 to 2006.10 In Fig. 3, a declining trend can be observed in overall per capita MSW discard
levels in recent years due to MSW policy interventions.
Considering the changing lifestyle in Taiwan, the Taiwan
Environmental Protection Administration (TEPA) and its
previous administrations have executed several countermeasures to control the rapidly growing solid waste discard
112
250,000
Amusement and
education
200,000
Transportation and
communication
Medicines and
medical care
150,000
Household
appliances
100,000
Housing
Clothing
50,000
Food
0
1984
Miscellaneous
1987
1990
1993
1996
1999
2002
2005
500
Miscellaneousincombustible
Miscellaneouscombustible
Moisture
400
Food
300
Glass
Metal
200
Plastics
Paper
100
0
1992
1994
1996
1998
2000
2002
2004
Number
Design capacity
(tonnes/year)
Throughput in 2006
(tonnes/year)
Landfill site
Incinerator
163
22
7 672 500a
864 354
4 163 968
113
100%
Others
80%
Dumping
60%
General
landfilling
Sanitary
landfilling
40%
Incineration
20%
0%
1990
1992
1994
1996
1998
2000
2002
2004
2006
114
Supplemental regulation
Waste Clean-up
Act
Effective date
Brief description
26/07/1974
31/07/1991
23/04/1997
17/04/2006
01/01/2007
11/12/2002
Environmental
Fundamental
Law
Resource Reuse
and Recycling
Act
03/07/2002
Restrictions of product
overpackaging
01/07/2006
Duration
Brief description
Stops the use of cargo for fixed point waste collection; instead, the
households have to directly dump their waste in the collection truck.
Bans temporary dumping at the collection sites
Builds an integrated recycling network among citizens (community), private
recycle companies (the recycling industry), local government, and the
recycling foundation.
Improves the business value of reuse, recycling, and treatment
Imposes restrictions on the utilization of packaging and tableware that are
composed of Styrofoam and plastics
Examines the previous solid waste generation trends and treatment capacity.
Proposes the zero-disposal society as the final goal of the solid waste
policy
Orders that households have to make a strict classification of waste before
collection by the local government.
Categorizes items that can be sorted, including the following:
(a) Combustible general waste
(b) Food waste
(c) Waste of potential resources
(d) Furniture and waste of large size
Recycles plastic bags with the MSW collection system in 14 major counties
Resource recycling
four-in-one project
01/07/2002 to date
Recycling of general
plastic bags
Recycling of waste
food oil
05/2006 to date
12/2004 to date
01/2005 to date
09/2007 to date
Includes waste food oil as one item on the mandatory recycle substance
list for specific food and service companies.
Encourages households to collect waste oil that is used in making food and
uses it as a raw material in the biofuel production industry
115
Target year
1996
2001
2006
2011
7
80
19.52
25
85
70
45
90a
80a
50
90
80
Denotes that the target value was achieved in the scheduled year
60
50
40
30
20
10
0
1990
1992
1994
1996
1998
2000
2002
2004
2006
Food
Housing
Clothing
Layer 2
...
Layer 3
Expenditure
Subcategories
MSW policy
variables
Socioeconomic
indicators
Expenditure
Categories
Layer 1
Layer 4
Food
waste
Paper
waste
Plastic
waste
Metal
waste
116
expenditure; vt denotes the column vectors of the explanatory variables that reflect the socioeconomic situation, e.g.,
the household attributes and the socioeconomic indices;
and bij is the column vector of the parameters.
Hov65rt and Savingt are used as the explanatory variables
in the case study for Taiwan in the MNL model. The parameters were estimated for the period from 1984 to 2003,
and the models were validated for the period from 2004 to
2005. Therefore, consumer behavior for the subcategories
of the consumption expenditure could be obtained.
ci ,t = ci ,t + i PConst c j ,t
j
and
ci ,t = bi + bik vk ,t
(1)
(2)
exp (vtT b ij )
cij ,t
=
cij,t exp (vtT bij )
j
(3)
where prob(yij,t) denotes the probability for spending consumption expenditure on the jth subcategory in the ith category in year t; j is the label of subcategories of consumption
Previous modules have established the quantitative relationship between the consumption expenditure and socioeconomic variables hierarchically. Thus, the estimated
consumption items serve as the input of the MSW discard
model.18 The MSW discard model is constructed by using
the SES model approach. Per capita MSW discard by waste
fraction is assumed to be a linear function of the consumption variables and MSW policy variables. The equations for
MSW fractions are formulated as follows:
WEm,t = a m + slT,t d l + zk ,tT z k
(4)
117
Table 7. Descriptions of variables used in the development of the MSW discard model
Variable
Description
Unit
WEi,t
WEm,t
kg
PConst
Fdt, AEt
FAEt
Fdj,t
HAj,t
MISt
Dum1
Dum2
Dum3
Incit
Recyt
%
%
Considering that policy measures are often launched 1 year earlier for some example areas, the setting of validating dummy variables are 1 year
before the formal year of introduction
(5)
where PConst is the vector of annual per capita consumption expenditure in year t, in 104 Taiwan dollar (NT$) at
2001 prices; PConst1 is the one-year lag term of PConst; f
() is the consumption function; and xt and vt denote the
vectors of macroeconomic variables and socioeconomic
variables in year t, respectively.
This study assumes a linear relationship between per
capita consumption expenditure and the explanatory variables. The parameters were estimated for the period from
1984 to 2003 and the model was validated for the period
from 2004 to 2005. Equation 6 presents the model fitting
results using the ordinary least-squares method (the value
in the parenthesis denotes the t statistics):
PConst = 0.209 PGDPt + 0.550 PConst -1
( 4.57 )
( 4.74 )
0.187 Unempt
( 2.35)
(6)
118
Table 8. Descriptions of the exogenous variables for the estimation model system of MSW discards in the Taiwanese case study
Indices
Socioeconomic variables
PGDPt
Unempt
Hov65rt
Savingt
MSW policy variables
Dum1
Dum2
Dum3
Incit
Recyt
Description
Unit
Dummy variable for the Resource Recycling Four-in-One Project: before 1997, the value
is zero, and is 1 otherwise
Dummy variable for the Restrictions on the Use of Plastics Bags: before 2001, the value
is zero, and is 1 otherwise
Dummy variable for the Mandatory Household Classification and Food Waste Recycling
imposed in Taipei city: before 2003, the value is zero, and is 1 otherwise
The fraction of MSW discard treated by incinerators in year t, a continuous variable
The recycled fraction of MSW generation in year t, a continuous variable
none
none
none
%
%
Scenario
A
BAU
low
high
4
2
5
21.63
25
20
4
5
2
The values are 9.95, 10.15, 10.35, 10.53, 10.60, 10.72 for the period from
2006 to 2011, respectively
At a constant increasing rate from 34.97 (actual value) in 2006 to 50
(target value) in 2011
82.76
The policy measures are activated in the three scenarios.
Consumption level
Growth rate of per capita GDP (%/year)
Saving, rate, Savingt (%)
Unempolyed rate, Unempt (%)
Elder population rate, Hov65rt (%)
40
Scenario A
Scenario B
30
Scenario C
True value
Model fitting
20
10
0
1984
1987
1990
1993
1996
1999
2002
2005
2008
2011
Fig. 7. Projection of annual per capita consumption expenditure. Scenario A, business as usual; scenario B, low consumption; scenario C,
high consumption
119
expenditure on food and its subcategories. Next, the projections of MSW discard levels for MSW fractions are estimated under the three scenarios (see Fig. 10) and depict the
possible evolution of MSW discard levels influenced by the
effects of the consumption factors and those of MSW policy
measures. In addition, the projection of annual per capita
MSW discard levels can be obtained by adding the projections of discard levels of the respective MSW fractions (see
Fig. 11).
From Fig. 7 to Fig. 11, the model fitting results for the
period 1984 (or 1985 in Fig. 7) to 2005 (or 2004 in Figs. 10
and 11) are good approximations, implying that the estimation model system has been successfully. The sequential
projections show that the different increasing or decreasing
trends of MSW fractions are caused by changes in the economic and social conditions. In Fig. 7, all the projections of
per capita consumption expenditure for the three scenarios
are increasing since the economy continues improving at
different speeds in the assumed scenarios. In Figs. 8 and. 9,
different socioeconomic variables in the assumed scenarios,
reflecting the influences of lifestyle changes on consumer
behavior, lead to different distributions of consumer expen-
10
Scenario A
Scenario B
Scenario C
True value
Model fitting
0
1984 1987 1990 1993 1996 1999 2002 2005 2008 2011
2.5
Annual per capita food expenditure by subcategory (104 NT$ at 2001 prices)
diture among the categories and subcategories. The different levels of consumption give rise to the corresponding
projections in the MSW discard level by waste fraction,
given the assumed MSW policy variables.
The evolution in terms of MSW discards by waste fraction is caused by a trade-off between the effects of the
consumption factors and those of the MSW policy measures. Therefore the discard levels of MSW fractions are all
the highest in the high consumption scenario (Scenario C),
intermediate in the business as usual scenario (Scenario A),
Milk
Vegetables-S
Vegetables
1.5
0.5
Beverages
Tobacco and betel nut-S
0
1984
1987
1990
1993
1996
1999
2002
3.5
ili
2.5
2
1.5
1
0.5
0
2006
2007
2008
2009
2005
2010
2011
120
100
80
100
80
60
60
40
40
20
20
0
1992
0
1995
1998
2001
2004
2007
2010
1992
1995
1998
2001
2004
2007
2010
2001
2004
2007
2010
2001
2004
2007
2010
350
100
300
80
60
200
150
40
100
20
50
0
0
1992
1995
1998
2001
2004
2007
20
15
15
10
10
1995
1998
2001
2004
2007
2010
50
40
1998
0
1992
1995
25
25
20
1992
2010
1992
1995
1998
12
10
30
6
20
4
10
0
1992
1995
1998
2001
2004
2007
2010
1992
1995
1998
2001
2004
2007
2010
Fig. 10. Projections of annual per capita MSW discards by waste fraction. Except for part d, the y-axis denotes annual per capita MSW discards
(dry-based kg/capita/year). For part d, the y-axis indicates the water contained in MSW discards (kg/capita/year)
121
700
Scenario A
600
Scenario B
500
Scenario C
True value
400
Model fitting
300
200
100
0
1992
1995
1998
2001
2004
2007
2010
122
Conclusion
2.
3.
4.
5.
6.
References
1. Taiwan Directorate-General of Budget, Account and Statistics,
Republic of China (TDGBAS) (2008) Online Information Service
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.