Interdisc伊Iillaryblformation Sciences， Vol. 14， No. 1， pp. 77-87 (2008)

Site Location of Projects in China by Clean Development Mechanism:

An Input Output Analysis*

Sungin NA'， Kiyoshi FUJIKAWA2， Tadashi HAYASHI3， Kazuhiro UETA4，

Yoshihisa INADA5， Akihisa MORI4， Takasei KUSUBE6 and Mitsuru SHIMODA7

'Hiroshima-Shudo Universitγ

2 Nagoya Universiη 3Takasaki Ciη UniversiりIof Economics

4 Kyoto Universiη 5 Konan University

61shikawa Pr，昭作ctureUniversiη 7 Applied Research lnstitute， inc.

Received October 24， 2006; final version accepted June 3， 2007

This paper examines which region in China would be suitable as a location for CDM projects using the Multi-Regional Input-Output Table China 2000. The criteria for selection used in this research inc!ude such socio-economic e仔ectsas output and employment increase induced by CDM projects， spillover effects of technology transferred by CDM projects from advanced countries and such environmental e仔ectsas reduction of C02 and S02 emissions by CDM projects. As a result of simulation analysis， the‘South coast'，‘North municipalities'， and the ‘Central region' were found to be potentia1 candidates as suitab1e regions for a site of CDM projects based on comprehensive evaluation of the above-mentioned criteria.

1. Introduction

Ueta et al. (2006) demonstrated that implementing Clean Development Mechanism (hereafter， CDM)' pr吋ectsin

China will bring various kinds of favorable e仔ects.Needless to say， CDM projects in China have large potentiality to

reduce emissions of greenhouse gas (hereafter， GHG). And they would assist China's sustainable development by

promoting job creation， equalizing income distribution or mitigating local air pollution and health damage in China.

Moreover， CDM projects in China could even assist sustainable development in the East Asian region as a whole by

mitigating trans-boundary acid thing deposition in the region.

Looking at CDM projects registered in United Nations Framework Convention for Climate Change (hereafter，

UNFCCC) by March 2007， China is the largest host country in terms of not only in the projects numbers but aJso in the expecting amount of Certified Emission Reductions (hereafter， CER). Besides， China has still large potentiality for

future CDM projects even though the transaction costs of CDM in China could be rather high. The pu中oseof this paper

then is to examine which region in China would be suitable as a location for CDM projects in case Japanese

organizations implement CDM projects in China.

Since the CDM scheme is a kind of foreign direct investment， one pu叩oseof CDM is to get CER by reducing GHG

emissions， but CDM has another important pu叩oseof supporting sustainable development in the host countries. ln

other words， socio-economic e釘ectson employment situation or income distribution are also important as to CDM

projects. Although a CDM contract itself is on a private basis， implementation of the pr吋ectrequires approval by the

govemments of both host and investing countries， as well as the CDM Executive Board of the United Nations.

Therefore， govemments and private stake-holders of both host and investing countries have to discuss various issues reJated to the proposed CDM projects.

Incidentally， one of the most difficult issues for a host country， especially such large country as China， would be a

location for CDM projects. While there are various points of view in regard to determining a location for CDM

projects， they can b巴broadlyclassified into two categories: e仔目的 inthe project-construction phase and those in the

project-operating phase. Generally speaking， socio-economic e仔ectsare emphasized in the project-construction phase，

and environmental improvement effects or resource saving e仔巴ctsare emphasized in the project-operation phase. As

for a measure of the socio-economic e仔ectsof CDM projects， the most well known would be the “output inducement

effect" or the “employment inducement e仔ect"calculated in the framework of input-output (hereafter， 1-0) analysis.

ln section 2， we outline the framework of 1-0 analysis and show an imaginary CDM project that this paper considers

taking China as an example of a host country. Naturally a host country tends to prefer a CDM project that has larger

economic inducement effects since such a CDM project is expected to contribute sustainable development in the host

• This research is白nancialIysupported by International Collaboration Project 2004/2005“日Istainableeconomic growth and structural reform in Japan" of ESRI (Economic and Social research institute) of the Japanese Cabinet Office. http://www.esri.go.jp/enJprj-2004_2005/menu-e.html

NA et al. 78

Correspondence of region and provinc巴.

A

B

C

D

E

F

G

Provinc唱 nam巴

I Heilongjia昭， 2 Jilin， 3 Liaoning I Beijing， 2 Tianjin I Hebei， 2 Shandong I Jiangsu， 2 Shanghai， 3 Zhejiang I Fりian，2 G叩昭dong，3 Hainan I Shanxi， 2 Hen直n，3 Anhui， 4 Jiangxi， 5 Hubei， 6Hunan I Inner-Mongoria， 2 Ningxia， 3 Shaanxi， 4 Gansu，

5 Qinghai， 6 Xinjiang I Guangxi， 2 Chongqin， 3 Guizhou， 4 Tibet， 5 Sichuan，

6 Yunnan

Table 1.

Region'name North East North Municipalities North Coast Central Coast South Coast Central Region

North West

South West H

Source: Okamoto and Ihara (eds.) (2004)， p. 5.

country. As indicated in OECD(2003)， we consider income distribution as an indicator of sustainable development. As

a proxy variable for equalization of income distribution， we used the standard deviation of region-wise output

inducement e仔'ectsof a CDM proj田t.On the other hand， such e町 ironmentalindicators as CO2 reduction and S02

reduction by a CDM project are also important in evaluating CDM projects. The estimation methods for such indicators

are also explained in section 2.

It is also natural that a host country's govemment would expect resource conservation in the country as a result of the

叩provalof a CDM project. In other words， the more resource conservation the CDM project can realize， the more preferable the project is for the host country's sustainable development. Then， in section 3 we introduce an indicator

called the “spillover rate of process innovation" as a measure of resource conservation in the project-operating phase.

This indicator is to measure how much an innovation can economize resource consumption in the other sectors than this

innovation actually occurred.

Moreover， one of the most urgent issues for China' sustainable development is mitigation of local air pollution that has recently been becoming more serious. Therefore， in section 4 we take environmental effects in the operating phase of CDM projects into consideration as another criterion to determine the location of CDM projects. And， in section 5， we wi1l show through a simulation analysis that the location of the projects would a百'ectthe total effects of CDM

proJects. We use the China Multi-Regional Input-Output Table 2000 (hereafter，“MRIO") compiled by the Institute of

Developing Economies (IDE， 2003) in Japan. In this 1-0 table， China is divided into eight sub-regions as shown in Table 1 and Figur巴 1.The great feature of these statistics is that the transactions of goods and services including not

only intra-regional intermediate inputs but also inter-regional ones are recorded by industry. Industry sectors in the

1-0 table we use are divided into 17 sectors as shown in Table 22.

Economic and Environmental E町ectsof CDM Projects in the Construction phase

Output inducement effects and C02 and S02 emission inducement e賀町ts

Leontief (1966) showed how input-output tables could be used to consider the relationship among industry-wise

demand and gross output. If column vectors of f and x respectively represent domestic final demand and domestic gross

output， the base equation of 1-0 analysis is as follows:

2.

2.1

、.，ノ11

(

Equation (1) shows how much of gross output (x) is required in order to satisfy the final demand (f). The element of

the matrix A， expressed as仇j，is the domestic input coefficient that shows the relationship of the domestic input of i

industry to the gross outputs of j industry and a matrix 1 stands for the identity matrix. The matrix of (1 -A)-. or B is

called ‘Leontief's inverse matrix' whose element， expressed as bij， means the induced output of j industry by one-unit

increases in the final demand for i industry3.

Furthermore， when we assume that employment vector e is proportional to the gross output， employment induced by final demand f can be expressed as in equation (2) where a diagonal matrix E is the matrix of employment coe伍cients.

e = E(I -A)一lf= E・B.f

x=(I-A)-lf=B.f

(2)

Based on this Leontief's inverse matrix， we examine the di釘'erenceof economic and environmental e仔'ectsof the

same size of CDM projects when the CDM projects are implemented in di任'erent陀 gions.

The output effect and employment effect of investments in CDM projects are estimated as follows

(3)

(4)

AXrニ B.Af"

Ae， ニ E.B.Afr，r=A，・・・ ，H

Site Location of Projects in China by Clean Developm巴ntMechanism: An lnput Output Analysis 79

Source: Institute ofDeveloping Economies (2003)， p.14.

Fig. 1. Correspondence of region ancl provinc巴.

where r repres巴nts巴achregion in th巴恥1RIO.Th巴 vectoraf，. m巴nasthe incr巴m巴ntof final demands by th巴 investm巴nt

for a CDM proj巴ctin r region4 and ax，. is th巴 V巴ctorof induc巴doutput by the CDM proj巴ctimpl巴m巴nt巴din r r巴gion，

and ae，. is th巴 vectorof induc巴demploym巴ntby the same CDM project implemented in r region

The巴lementsof vector Af，. correspond to construction of coal-burning power plants with a total nominal capacity of

3，000 MW as shown in Table 3. The industrial component of th巴 investmentamount was estimated based on the

Japan巴secas巴carn巴din th巴Fix巴dCapital Formation Table for Inpllt-Output table in Japan， 20005 Th巴mainindustri巴s

whos巴goodsar巴demandedby CDM projects are 'El巴ctricmachinerγand ‘Construction' and the demand for these two

industri巴sexplains approximately two thirds of the total investment amount.

Next， based on the v巴ctorax，. of Olltput inducement， we can estimate induced CO2 and S02巴missionsas follows:

企c，.= CdX，.，

As，. = Sdx，.，

(5)

(6)

wh巴rethe vectors of Ac，. and As，. respectiv巴lystand for the ernissions of CO2 and S02 and C and S stand for a diagonal

matrix of emission co巴飴 ci巴ntsof S02 and CO2， respectively. Emission coefficient matr・icesC and S were estimated bas巴don th巴 followingstatistical data6.

• En巴rgyconsumption by region and industrial s巴ctor:China Energy Statistics Yearbook 2000-2002. Index巴Sof

output values and output amounts: China Statistica! Yearbook 2001

• Conversion factors of calori巴s:China En巴rgyStatistics Yearbook 2000-2002.

• CO2 emission co巴fficient:National Institute of Science and Technology Policy of Japan (巴d.)(1992).

• S02 emission coefficient: National Institute of Sci巴nc巴 andT巴chnologyPolicy of Japan (巴d.)(1992).

For CO2 emissions we assumed the sam巴 emissionco巴fficientfor all regions， while with r巴gardto S02 emissions

we needed to b巴 carefulsince the slllfur content rates of coal aI巴 differentamong different regions in China7

The differences of sulfur cont巴ntrates wer巴 adjllstedby r巴gionalS02巴missioncoefiicients. Moreover，出ose

co巴fficientsdo not consider waste-gas d巴sulfurization，so emissions of S02 considered in this res巴archar巴 not‘r巴al巴missions'but ‘pot巴ntial巴missions'8.

2.2 Economic and environmental characteristics by region in 2000

In this section， we revi巴w the economic and巴nVlronm巴ntalcharact巴risticsin China by r巴gion.Tabl巴 4shows

NA et al. 80

Basic se芯torc1assificatiolI AgriclIlture Coal min川 gand Processing Crlldc petrolellm and natural gas prodllcts Metal orc mining Non-ferrolls mineral mining Manllfacture of food and tobacco processing Textile goodぉ

Wearing apparel. leather， furs， down and relatcd prodllcts Sawmills and furniture Paper and prodllcts， printing and r官cordmedilll11

I官production

Petroleum processing and coking Chemicals Nonll1巴talminer百1products

China's sector classilication. Table 2.

l一2

3

4

5一6ヲ'8

17Sect()rs Agricultul巳

Mining

Food product討

2

3

9

10

Textile and wearing apparel

Wooden products

4

5

Paper and printing 6

日ロ一日

Chemical products 7

Metals smelting and pressing 恥1etalproducts Machinery and eqllipment Transport equlpment Electric equipment and machinery El巴ctricand telecommunication巴quipment

Instrllments， m巴ters，cultural and office machinery Maintenancc and r官pairof machine and cqllipment Othcr manut'acturing products Scrap and w出 teElectricity唱 steamand hot water product旧nGas production and supply Water prodllction and supply

Constructlon Transport and warehousing Wholesale and retail trade Servic巴討

は凶一凶一げ一同日山口一羽引

nn一MA

おお一幻一お汐一知

Non-metal mineral products 8

Metal prodllcts 9

Machinery Transpo口equipment

σb

n

l

-c

l

長l

n

a

mm

r

1

ρiv?引uk

出向山

omy

Electronic prodllcts

Electricity， gas and watcr supply

10 11

12

13

14

Construction 15

Trade and transport 16

Se円 lces17

Source: Institute of Developing Economies (2003)， p. 24

the region-wise gross output (proxy variable of income) and region-wise巳missionsof CO2 and S02 in the y巴ar2000.

The national total gross output is 19，984 billion RMB. The ‘Central Coast'，‘South Coast' and ‘N orth Coast' account

for 23.1%，14.99もand14.3% respectively， so that the total of these thr巴ecostal regions account for more than half

of the national total outpue. The economic size of the ‘C巴ntralRegion司 isalso large at 17.5%.

According to the recent China's rapid economic growth， China's energy consumption also increased rapidly and the

total emissions of CO2 and S02 respectively became as much as 944.1 million ton-C and 9.2 million ton in 2000. As for

th巴regionaldistribution of CO2 emissions， the ‘Central Region'，‘Central Coast' and ‘North Coast' accounts for 32.7%，

14.2%， and 12.9% respectively. Among them， emissions in the ‘Central Region' are extrem巴Iylarge. On the other hand，

regarding S02 emissions， the ‘Central region'，‘Central Coast' and ‘North Coast' account for 21.6%司 16.1%，and 15.69も

respectiv巴Iy.The reason why the regional distribution of C02 and S02 emissions is not same as that of the gross output is that energy consumption pattern and indllstrial structures are not conformable among regions. Next， let us look at the

emission coefficients. The national average of C02 emission coefficients (national total C02 emissions j national total

output in China) is 47.2 ton-Cjmillion RMB and the regions where CO2 coefficients exceeded the national average are

such four regions as‘Central Region'，‘North West'，‘North Municipalities'， and ‘North East'. In other words‘carbon

intensity in thes巴regionsis larger than that in other regions. On the other hand， the national average of S02 emission

coefficients is 0.46 tonjmillion RMB and the regions whose S02 coefficients exceeded the national average are such five regions as‘No口hWest' ，‘NoロhMunicipalities'，・NorthEast' ， 'Central Region' and ‘North Coast'. In these

regions， the share of coal is relatively large in the energy consumption besides the白ulfurcontent in the conslImed coal

is relatively high.

Simulation analysis on Economic and Environmental Effects of CDM investment in China's power sector

in the Construction Phase 2.3

On th巴basisof such regional characteristics‘let us simulate the economic and environmental e仔ectswhen the CDM

project is implemented in a certain region in 2000. The CDM investment is assumed to be implemented as shown in

Table 3. Table 5 shows the results that we obtained in eight simulation cases. For example， the first row shows the case

l¥1

Site Location of Projects in China by Clean Development Mechanism: An lnput Output Ana¥yS1S

Table 3. lnvestment amounts of CDM projects (3，000 MW) by industry

(MilIion RMB). 伊

川

-

可抗白t

噌

開一

0

0

0

1

D

0・nMD3239rbρ88ゆ一

nM

m一な

0

0

0

1

0

0

0

0辺

U

4

0

0

6

7

0一路

-

4

L

A

U

1

ζ

J

，、d'

l

'

a

‘

s

-

O

A

U

弓

4

・1aA『弓

f

pi--

V

-

'

l

'

l

今

3

n一官----

r一一Sectors Agriculture Mining Food Textile and wearing apparel Wooden products Paper and printing Chemical products Non-metallic mineral products Metal products Machinery Transport equipments Electric machinery Other manufacturing products Electricity， gas and water supply Construction Trade and transportation Services Total

12345678901234567

1

1

1

1

1

1

1

同町(ωn/miIl. RMB)

0.63 0.64

0.50 0.32

0.25 0.57 0.78 0.42

0.46

明ぷえ4

Share

Output and emissions of CO2 and S02 in 2000.

C02

Table 4.

Output Million RMB

13.3% 6.6% 15.6% 16.1% 8.0% 21.6% 9.5% 9.3%

ω0.0%

10ooton

1223.7 607.0 1432.1 1475.2

733.9 1981.1 875.4 848.9 9177.3

In函両一(ton-Cfmill. RMB)

54.08 54.17 42.78 28.96 21.35 88.21 69.06 40.71 47.24

Share

ゆ一川一明一間一同一向一間一同一山

Mill. ton-C 105.3 51.0

121.9

133.9 63.5 308.5 77.5 82.4 944.1

Share

9.7% 4.7% 14.3% 23.1% 14.9% 17.5% 5.6% 10.1% 100.0%

1，947，259 941，944

2，849，491 4，625，353 2，975，666 3，497，790 1，122，470 2，024，449 19，984，423

Region

North East North Municipa1ities North Coast Central Coast South Coast Central Region North West South West Tota1

A

一B一C一D一E一F一G一H

where CDM investments are implemented in the North East region and the second row shows the case where CDM investments are implemented in the North Municipalities， and etc.

The first column shows the output inducement effects of CDM projects. We can see that the CDM projects bring relatively large economic e百'ectswhen the projects are implemented in the ‘Central Region'，‘North Coast' and ‘North

East'. The second column is the standard deviation of output inducement effects for each region. ln this paper， standard deviation of the increases in regional gross outputs induced by CDM pr吋ectsis regarded as a proxy variable of income equalization e仔'ect.The standard deviation of induced regional gross outputs is relatively large in the case where CDM investments are implemented in such regions as‘Central Region'，‘North Coast' and ‘North East'. These regions. therefore， are not necessarily suitable as sites for CDM projects if‘income equalization' is a factor of high importance for the Chinese govemment as CDM approval criteria. The third column shows the employment e仔ectsof CDM

projects. Since westem rural areas such as the ‘North West' and ‘South West' are relatively labor intensive regions，

CDM implementations in these areas bring relatively large induced e仔'ectsin employment. The employment e紅白tfor the ‘Central Region' is also considerably large.

The fourth and fifth columns show increases in emissions of C02 and S02 in the construction phase of CDM projects. It is in the ‘Central Region' that has the largest effect on CO2 emissions followed by the ‘North West' and

‘North East'， while the ‘North West' has the largest increase in S02 emissions followed by the ‘North East' and ‘Central Region'.

Spillover Ratio of Innovation

Process innovation and product innovation10

CDM is a transplantation of new technologies from developed countries to developing countries. Innovation effects are reftected by changes in the input coe伍cientsin the framework of 1-0 analysis. There are two types of innovation; process innovation and product innovation.

3.

3.1

NA ef al.

S02

1000ton 6.70 4.94 5.79 4.42

2.48

6.40

6.77 4.96

C02

Mill‘ton-C 0.62 0.45

0.53 0.45

0.26 1.00 0.63 0.50

Effects of CDM pr'Ojects in th氾 constructionphase.

Output Emploヴment

5tandard dほviation 1000 persons 2673.0 250.2 1613.1 100.1 2741.5 299.0 2295.7 191.7 1511.4 151.4 2656.5 492.3 2098.9 406.4 2490.2 566.0

Table 5.

E五tputTOta!'

-li枠 MillionRMB10，140.1 6，364.8 10，264.8 9，368.2 6.297.7 10，418.0 9，475.7 9.755.1

h伊 E

North East

North Municipalities

North Coast

Centra¥ Coast

South Coast

Central Region

North West

South West

o.G

A

一B一C一D一E一F一G一H

• Process innovation: efficiency improvements in an industry. More output can be produced with the same amount of inputs. Process innovation implies a shift in the production function and the isoquant. Hence， the coefficients in

the column are changed. . Product innovation: efficiency improvements of inputs. In each of n production processes， the same amount of

output can be obtained with a smaller amount of the product as an input. Hence， the coefficients in the row are

changed. Typical CDM investment considered in this research is， for example， improvement of production process for overall

energy saving. Innovation with CDM project， therefore， is regarded as a process innovation. We focus on process

innovation in this paper.

SpiIlover coefficients of process innovation

Innovation in industry k is defined as shown in the following equation. The parameterαstands for the rate of

innovation.

3.2

(7a)

(7b)

百ik= (1 -α)aik (0く αく 1， i = 1，'・・ ，n)

瓦ij= aij (i = 1，'・・ ，nand j手k)

Supposing that A and瓦respectivelystands for an original input coefficient matrix and the one with improved

efficiency司 thenthe process innovation is expressed as follows.

(8)

The vector ek is a unit column vector where element k is one. Supply~emand balance after process innovation is expressed as follows. B is an inverse matrix of Leontief co汀espondsto A.

瓦 =A-α(Aek)e~

(9)

CDM projects correspond to process innovation.The more resource consumption a CDM project can economlZE

in the entire economy， the more e百'ectivethe CDM is in the context of environmental conservation. The spi110ver eぽECtsof process innovation SK is deaned as the ratio of the output change (deemse)伽 toccurs m secωrs ot田 rthan the innovated industry k in the whole output change (decrease)in the economy-In shod，this is an indicator to see how much innovation in industry k can economize outputs in industries other than industry k.

支=Bf

乞(王i-Xi)

i手kSk =ミ=有一一一一一一一

) '(主i-Xi)

Using the formula for the inverse of a sum of matrices 11 the difference of ex ante and ex post in Leontief' s invers巴

matrix yields the following equation.

、‘aF

AU

l

，SE

、

、••

F

1A

l

J

，.‘、B-B= 一 αB(Aek)e~B1 +αekBAek ... A

、‘，F

F l

l

(

This equation can be further transformed since BA = B -1.

亘-B=一 α(B-I)eke~B 1+α叫(B-I)ek

The denominator of the right hand side is a scalar of 1 +α(bkk一 1)which shall be denoted as I1k = 1 +α(b紘一 1)・

And since bij -bij = e;(B -B)ej， each element of豆一 Bis written as follows.

Site Location of Projects in China by C1ean Development Mechanism: An lnput OUlpUI AnalYSlS

五リ -bij二 αbikbkj/llk

Eり-bkj =一α(bkk一 1)bkj /ηk

Let us compare exαnte output x = Ax十 fand ex post output玄=A支+f.

奇心 =αbik~二 bkj抑止

Xk - Xk =一α(bkk- 1) I:>kμ l

83

(¥2a)

(12b)

(13a)

(13b)

The spillov巴reffect of process innovation i日definedas the ratio of the outpul change (decrease) that occurs in sectors

other than the inr叩 vat印edindus幻t町 kin the total output change (decrease) in the巴c∞onomy.Let us defi祉r問 thesum of

column k of Leontief' s inverse matrix B as Ck・Then，the spil10ver coefficient of process innovation Sk is defined as

follows.

PEt-xt)bk

YK=ZFZ7=τ7了

Note that the coefficient Sk is independent of final demand jj or the innovation rateα.

3.3 Spillover rate of process innovation in China

( 14)

As equation(14)shows，Sk depends 0111wo factors;the overall backward linkage effect ck and the feedbackeffECt to the own sector hkk. As 川 heformer fac肌耐largeris its overall backward linkage e紅白t，白 largerthe spillover凶 e

becomes. As to the 1atter factor， the larger is the industry' s own input， the smal1er the spillover rate becomes. Table 6 shows the top 40 in 136(= 17 * 8)‘industries*fegions' whose spillover巴ffectis large. The most conspicuous

sector is ‘construction' whose spillover e汀ectfor all regions is considerably large. This is because of the industrial

featur巴 of‘construction'that the backward linkage e仔ectis large but the feedback from the own industry is smal!.

Another notab1e sector would be‘Electricity. gas， & water'・Itis also imaginable that th巴own-inputin the ‘Electricity，

gas， & water' industry is so small that the spillover effect is large in comparison with other manufacturing sectors.

As is shown in Table 6， CDM projects in the North Municipalities or North Coast are preferable from the point

of view of spillover effects suppose‘Electricity， gas‘& water' is a target of CDM projects in China.

4. Net Environment Effects of CDM Projects

4.1 Environment effects of CDM projects in the operating phase

A CDM project has two di汀erente釘ects;one is an economic 巴仔'ectin the construction phase and the other is

an environmental e仔'ectthrough the resource saving functions of CDM plants/facilities. However， it is important to

note that CDM projects have negative environment impacts in the construction phase， since the CDM investments

consume additional natural resources including energy. Therefor巴， we need to offset these e仔'ectsin order to evaluate

the overall e仔巴ctsof a CDM project. As is well known， China's main source of SOz emissions is coal consumption and the largest coal consumer in China

is the electric power generator. Japan's power generation technology is well-known for its high efficiency of energy

conversion. Therefore， it is quite plausible that CDM projects are implemented in the power generating industry and

technology is transferred from Japan to China. The average of annual increment of electric power capacity in China

around the year 2000 was approximately 20，000 MW12目 Andwe assume in this paper that coal-burning power plants

whose nominal total capacity is 3000 MW， which is 15% of total increment， would be replaced as a CDM project in the

year 2000. According to a power plant specialist in a Japanese paper generator， the e白iciencyof power generation in Japanese is

20 to 30% higher than that in China. We， therefore， assumed the efficiency improvements by CDM would be 25%.

Therefore， the input coefficients of intermediate inputs in the electricity industry in the host region would decrease by

this improvement rate. However‘ the CDM project we consider does not replace the whole capacity of power generation in the co汀espondingregion. Then， we need to adjust the efficiency improvement rate by the relative scale of the CDM project13.

Table 7 shows the‘adjusted rate of improvement' in each region. Though the ‘North Municipalities‘is an exception，

the improv巴mentratio is in the range of 1 to 2% reflecting the capacity of power generation of each region.

We assume the operating period would b巴tenyears from 2001 to 2010. Ueta et al. (2005) discusses the macro effects

of a CDM pr吋ectin China and presents a macro projection for the Chinese economy. We used this estimation as the baseline forecast for a series of real tinal demand in the r region， f， 14. Supposing that A is an original input coefficient

matrix and A is an input coe拍cientmatrix with improved efficiency， then economized total output dXr is expressed as follows:

84 NA et al.

Table 6. Spillover rate: Top 40 in 136(= 17 x 8)‘lndustri巴sx Regions'.

公 Region Industry Spillover

AOl5 Construction 0.9993

2 FOl5 Construction 0.9969

3 0015 Construction 0.9968

4 BOl5 Construction 0.9956

5 GOl5 Constructio日 0.9954

6 EOl5 Construction 。目9946

7 COl5 Construction 0.9932

8 HOl5 Construction 0.9929

9 EOIO Machinery 0.9893

10 B002 Mining 0.9893

11 GOl3 Other manufacturing products 0.9810

12 0002 Mining 0.9794

13 E002 Mining 0，9757

14 GOll Transport equipment 0，9738

15 AOl3 Other manufacturing products 0.9658

16 COl3 Other manufacturing products 0.9652

17 BOIO Machinery 0.9650

18 BOl4 Electricity & gas & water supply 0.9645

19 B005 Wooden products 0.9616

20 GOIO Machinery 0.9613

21 C014 Electricity & gas & wat巴rsupply 0.9600

22 F012 Electronic products 0.9583

23 B008 Non meta1 mineral products 0.9576

24 HOl3 Other manufacturing products 0.9572

25 COll Transport equipment 0.9557

26 GOl4 Electricity & gas & water supply 0.9537

27 AOl4 Electricity & gas & water supp1y 0.9492

28 COl2 Electricity & gas & water supply 0.9471

29 G005 Wooden products 0.9466

30 0014 Electricity & Gas & Water 0.9461

31 BOl3 Other manufacturing products 0.9445

32 HOO8 Non metal mineral products 0.9445

33 HOIO Machinery 0.9425

34 H014 Electricity & gas & water supp1y 0.9411

35 0013 Other manufacturing products 0.9390

36 G009 Metal productぉ 0.9280

37 G007 Chemica1 products 0.9268

38 C008 Non metal mineral products 0.9252

39 AOl7 Servic巴s 0.9213

40 F014 Electricity & gas & water supply 0.9184

A North East E South Coast

B North Municipalities F Centra1 Region

C No口hCoast G North West

D Central Coast H South West

-1 iT A¥-1 dXr = Xr -Xr = [(1 -A) ー (1-A)-l]fr. (15)

Assuming the symbol with the upper bar means that based on CDM technology， economized COz and SOz emissions

are also defined as follows.

dCr二 Cr-Cr = C(Xrー支r)= C. dXr

dSr = Sr -s，: = S(Xr 支r)= S. dXr

(16)

(17)

If we repeat this calculation for the decade of 200ト2010and sum up all of the results， th巴environmentaleff巴ctsof

CDM projects in the operating phase wiJl be calculat巴das follows:

仙 r= L dCrt (18)

85

Adjusted巴日iciencyimprovement by regions.

Adjusted rate of

efficiency improvement

ト97%

7.65%

1.489も

1.31 %

l目57%

0.98%

1.93%

1.64%

Site Location of Projects in China by Clean Developmenl Mechanism: An Inpul OUlput Analysis

Table 7.

Region

No口hEast North Municipalities

North Coast

Central Coast

South CoaSI Central Region No同hWest South West

A一B一C一D一E一F-G-H

Environmental effects of CDM projects in the construction phase， operating phase and net e仔ect.

Construction Operation Net， C02 S02 C02 S02 Mill. ._. MilL

1000ton ton-C . ~.----~.. toトC

0.62 6.70 -2.62 0.45 4.94 -2.87

0.53 5.79 -2.19

0.45 4.42 -3.94 0.26 2.48 -4.74

1 .00 6.40 -5.17

0.63 6.77 -1.38

0.50 4.96 ー1.42

Table 8.

3仔ect

S02

n

o

I

AU

)

(

)

{

l 25.49

32.69

21.86 40.47

55.41

40.83

9.51

11.66

問一肌叫一川一川両一川一州一川一山一肌

n

o

t

nu (

l 32.19

37.64

27.65

44.89

57.89

47.22

16.28

16.62

North East North Municipalities

North Coast

Central Coast

South Coast Central Region

North West

South West

A

一B-C一D一E一F一G一H

Remark: net effect is not necessarily consist巴ntbecause 01' rounding e町or.

」U4Z4

(19)

The first column of Table 8 shows the environmental e仔'ectsof the construction phase which is the same content

as shown in Table 5， and the second column of Table 8 shows the environmental eff削 sin the operating ph蹴 of

the CDM proJEU-Concerning C02mitigation eiTectsin the operating phase，the region that records the largest effect is the ‘Central Region' followed by ‘South Coast' and ‘Central Coast'. Concerning S02 reduction， reduction in the ‘South Coast' is the largest， followed by ‘Central Region' and ‘Central Coast司.

tds，. =

Net environmental effects of CDM projects

As mentioned above，CDM investment brings an increase in Emissions of C02and S02in the construction phase，

although they will decreas巳 inth巴 operatingphase. Therefore， we need to know the net reduction of CO2 and SO今

emissions in 0地 rto evaluate the environmental巴仇ctsof the CDM p叫 ect.That is to s祢 thenet dect ofCDM li

reductions of CO2 and S02 respectively are defined as“(18) minus (5)" and “(19) minus (6)".

4.2

(20)

(21 )

nc，. = tdcr -ACr

nSr = tdsr - As，

The net e仔'ectsby region defined by (19) and (20) are shown in the two right most columns of Table 815. It is

;l;;it;;札口;eZ2222113JeU2122d:7311iztrz:Ji込官出::;iCoast' and ‘North Municipalities'. And the ‘South Coast'， the region with the largest CO2 reduction， was able to reduce

4.48 Mill. ton-c， approximately six times as much as th巴reductionof 0.74 Mill. ton-C in the‘North West'司 theregion

WIth the smallest C02rcdu?tion-Similar results are observed concerning S02reduction.The ‘South Coast'. the region

with the la伊 ?tS02reduct10IL was able to reduce 55.4l kilo-ton-approximately six times as much as the reduction of 9.51 kilo-ton in th巴‘NorthWest'， the region with the small巴stS02 reduction.

It ls one of the katures of our study that the C02reduction inEach region renects theinput structure of thepower

genem!m industry in thE region.In other words.even if the scales or wes of CDM p吋ectsare same， CER

叫 UIMmsared!釘erentamo中旬ionsbωuse of the di能 re附 inthe location of tlば DMproj配 ts.In summ町， the

simulation analysls suggests that the location ofCDM projects should bedetermined cautious旬、 takinginto account not

only CER acquisition but also thesize of the ripple effect the target industry has on other industries and Other regions.

NA et al. 86

Rank of the site location in CDM.

Principal compon巴ntVector Rank

-0.976 5

1543 2

-1.525 6 0.673 3 4.093 1

-0106 4 1582 7

-2.119 8

Runk 6

2

5

4

2

7

Point system Total point

29

26 28

27 26 20

30

30

Table 9.

7

Region

North East North Municipalities North Couは

Ce泊tralCoast

South Coast Central Region

North West

South West

A

一B一C一D一E一F一G一H

Concluding remarks

This paper considers which region is preferab1e as a 10cation of CDM projects from the various points of view of

socio-economic and environmental e仔ects.Let us summarize several facts we observed in this paper.

• The 'Central Region' is preferred based on output increases in the CDM construction phase.

• The ‘South Coast' is preferred based on the standard deviation of output increases in the CDM construction phase. ・The‘South West' is prefe汀 edbased on the employment increases in the CDM construction phase.

・The・NorthMunicipa1ities' is prefe汀edaccording to the technological spillover criterion.

• The ‘South Coast' is prefe打 edbased on the net effects of the reduction of CO2 and S02巴missions.

These results suggest that it is very diffic山 todetermine the most suit油 le10cation for a CDM proj巴ct.In other

words， it is difficu1t to r巳concilethe profit-Ioss re1ationships among stake-holders concerning a CDM project.

However， we here attempt to compile these results into one variable to rank the regions in order. We apply two

methods in order to make a rank order. The resu1ts are shown in Table 9. The first method is a‘point system'， where

each region is given a point of 8 to 1 based on the score for each criterion introduced in this paper16. This method

eventually gives the same weight to all of the eight criteria to int巳gratethe information. Judging from the overall score

of thes巴 criteria，the most preferable region as a 10cation for CDM is the ‘Central Region' followed by the ‘South

Coast‘and ‘North Municipa1ities'. The ‘Central Coast' is fourth. Another method is an app1ication of 'principal component analysis'. It is a multivariate analysis that is used to

consolidate differ巴ntkinds of information 17. More specifically， principal component ana1ysis finds the vector that has

the maximum varia即巴 amongany linear c∞ombi削悶

ana1ysis， the prefe汀edregion as a location for CDM is the 'South Coast' followed by the‘North Municipa1ities'，

‘Centra1 Coast' and ‘Central Region'. Ev巴nthough the two resu1ts are not necessari1y identica1， it turns out that the ‘South Coast'，‘判No凹rt山hMu叩I汀1mκ凶ci増pa叫li凶t1e目s'、

and 'Ce臼e釘nt凶 Regiぬon'，ar，陀巴 can吋di凶da剖te邸sfor the most s乱叩山u凶lItawe cited in t出hiおspape訂r.Thatis to say，that while china's government expects that CDM would be one measure to reallZE

‘G削 westernd州 opment'，ou山rsimu山I

in c∞omp帥 ensl竹V均 mCooP 巴ration of the inv 巴s託吋tingand host c 【ou山mtri巴s is indispensable tωo the impμ1ementation of a CDM project. However，

m o s t of m e a r巾 s on CDM see 山 effe ω ofCDM from 山 V叶i巴wp仰阿Oω1n川山t0ぱf附 Anr肌 Ic∞our川Iwhi1e t出h巴yrarely巴mphas幻iz巴 theviewpoints of d巴velopi凶ngcountries. Therefore， we believe that this paper is unique

because this paper provides profitable information for the host country. Since the Kyoto protocol came into force on 16February2005，AI111cx I countries haVE to achieve their reduction

targets in the commitment period2008t020 12.However，it is Haid that the domesticmeasure8would not be enough

for Japan to achieve the targ巴tof Kyoto Protocol;thErefore it would be inevitable for Japan to make use of .the

Kyoto mechaIMEIn-We also believe that the information in this paper would be useful whm Japanese arms negotlate

with the Chinese government.

5.

Notes 1 C1ean dev巴10吋 Tlentmechanism (CDM)is one of thenexibility mechanisms in Kyoto Protocol-And CDM has

such問州巴ctivesas e同cientreduction of Gr削 1附 eGas (GHG) emissions and contribution to the

sustainable develc吋 n巴ntof host countries.

2 :口;:: :訪U:町:;主t註叫Lz昆出:2古;2;お拡杭r;芯訳訂:刃r灯:f点η北巾巾f:rt官叩t;2r日口;2r出Jy2i;l;ば:

3立:110凶 ewe used is of 17間仰sw抽 8時 i肌批 S四川einput coe伯C削 matrixand the

Leontief s inverse matrix is 136 by 136.

Site Location of Projects in China by Clean Developlllent Mα:hanism: An Input Output Analysis 87

4 For exampl巳 rmeans the ‘North West'， the‘North West' part of the vector Afr contains positive numbers and the

other elements are zero.

ラ Asupplem巳nttable of the Ministry of Internal Affairs and Communications of Japan (2004).

6 For details， see Ueta et al. (2005). 7 AII the coals dug out in a certain region， 01' COllrse， are not necessarily consumed in th巳 sameregion. However.

ther巴isno data showing that each power plant lIses what kind of coals. And it would be natural that most of coals

consumed in a power plant are from neighboring region taking the distribution cost of coals into accollnt. We，

therefore， used the sulfur content rates of the coals in the same I巴gionas the power plant exists as a

approxlmatlon.

8 There is no reliable statistics concerning operation of desulfurization equipm巴ntthough a few power plants in

China recently began to introdllce desulfllrization equipm巴nt.In addition， it is said that desulfurizatIon equipment

scarcely operates even if it is set up.

9 This concentration of economic power is a reHection of a policy of foreign capital introduction to the coastal

reglOn.

10 Description in this section is based on Dietzenbacher (2000).

11 See Henderson and Searle (1981).

12 See Hama (2006). Also， according to Xing Ying (2004)， the government plan for annllal increment of electric

power capacity up to 2010 in China is approximately 30，000MW.

13 For example， the capacity of the power plant invested as CDM is 10% of total capacity in the corresponding

region， the adjusted efficient improvement rate is 2.5% (25% times 0.1).

14 According to the prediction of Ueta et al. (2005)， the growth rate of real private consumption， real investment，

real export， and real GDP for 200-2010 would be 6.2弘 6.5%，9.1 % and 6.8%， respectively. We assumed that

final demands of all the regions would increase identically at those growth rates.

15 In the meantime， estimated C02 reduction in our research would be larger than CER obtained by the current rule

for CDM. This is because our research considers the overall effects including ripple effects to other industries，

while CER obtained by CDM focuses on only the direct effects in the establishment of the actual CDM project.

16 The r巴gionof the first place gets 8 points and the region of the last place gets 1 point.

17 Se巴 forexample Theil (1971) section 1.9. In this paper， we used vectors after normalization.

18 This is a tentative concJusion that can differ according to the criteria taken into account.

19 The countries that signed Annex 1 of the United Nations Framework Convention on Climate Change (UNFCCC)

are called Annex 1 parties or Ann巴x1 countries. Ann巳x1 parties inc1ud巴suchindustrialized countries as OECD

members as of 1992， the Russian Federation， the Baltic States， and several Central and Eastem European States.

Kyoto protocol signed in the third conference of UNFCCC (COP3) stipulates that Annex 1 countries should

reduce greenhouse gas (GHG) emissions by 5% comparing with 1990 emission level.

REFERENCES

Di巳tzenbacher，E. (2000)，“Spillovers of Innovation Effects，" Journal of Policy Modeling， Yol. 22， No. 1， pp. 27-42.

Hama， K. (2006)，“China en巴rgycrisis brought change 1"01巴nergyconservation development strategy，" Soka University Bulletin of Chinese Studies， Soka University， Yol. 9， pp. 1-47 (in Japanese)

Henderson， H. Y.， and Searle， S. R. (1981)，“On Deriving the Inverse of a Sum of Matrices，" Siall1 Review， SocielツforIndllstrial αnd Applied Mathematics， Yol. 23， No. 1， pp. 53-60.

Institute of Developing Economies (2003)， Multi-Regiollallllpllt-Output for China 2000， Statistical Data Series No. 86， Institute of Developing Economies， Tokyo.

Leontief， W. (1966)， Inpw-Output Economics， Oxford Univer日ityPress， New York.

Ministry of Internal A白'airsand CommunicationぉofJapan (2004) lnplll-Olllput tahle in Japan， 2000.

National Bureau of Statistics of China (2001). China StatistiαIi Yearbook 2001， China Statistics Press.

National Bureall of Statistics (Department of Industry and Transport Statistics) and National Development and Reform Commission of China (eds.) (2005)， China Energy Statistical Yearhook 2000-2002， China Statistics Press.

National Institute of Science and Technology Policy (巳d.)(1992)， The Global Environmenl and Energy Use in Asia， Ministry of Finance Pr巴ss，Tokyo. (in Japan巴se).

OECD， 2003. OECD Environmental Indicators: Deνelopmenl， Measurement (/nd U.白ぽ. Downloadable at http://www.oecd.org/ 巴nvlindicators.html.

Okamoto and Ihara (eds.) (2004)，“Spatial structure and regional dev巴lopmentin China: International input-output approach，" IDE development perspeιtive series No. 5， Institute 01' Developing Economies， Tokyo.

Theil， H. (1971)， Principles of ecollometrics， Wiley， New York.

Ueta et al. (2005)，“Win-win strategy for Japan and China in climate change policy." Downloadable at http://www.esri.go.jp/ jp/pη2004_2005/kankyou/kankyou 16/09-2-R.pdf

Ueta et al. (2006)，“Comprehensive Analysis and Evaluation of CDM Projects in China." Downloadable at http://www.esri.go.jp/jp/prj-2004_2005/kankyou/kankyou 17 /04-1-R-I.pdf.

Xing， Y. (2004)，“present sitllation and future outlook 01' China's electricity demand and sllpply" Overseas Electric Power，

Japan Electric Power Information Cent巴r，S巴ptember，pp. 34-39. (in Japanese).