第2章 CGEモデル - 内閣府経済社会総合研究所

第２章
ＣＧＥモデル
１. 研究要旨
１.１
はじめに
「経済・環境の相互作用の総合的分析」のための分析モデルとしてのひとつの候補と
しての CGE モデルの研究計画について述べる。
CGE と い う 用 語 に つ い て の 注 釈 か ら 始 め る 。 CGE （ Computable General
Equilibrium）と AGE （Applied General Equilibrium）とは、同義であるかのように
使われることが多い。しかし、それらは、別の研究系譜を示す呼称であると考えた方が
よい。
1950 年代に一般均衡の存在証明がアロー、ドゥブルー、マッケンジー、二階堂らに
よって盛んに研究されたことは周知の通りであるが、それは一般均衡がどこに存在する
かをつきとめようとするものではなかった。しかし、一般均衡の位置を具体的に計算で
きれば、政策分析等への幅広い応用の可能性が期待できる。実際、スカーフとその研究
を受け継ぐひとびとは、不動点アルゴリズムの開発・改良に取り組みながら財政政策へ
の応用を中心に多くの成果をあげてきた。この研究系譜は AGE と呼ばれる。
それに対して、世界銀行等の国際機関を中心に開発政策の形成・分析の用具として発
展してきたもうひとつの研究系譜がある。それが（AGE とは区別された意味での）CGE
である。それは、開発政策の分析用具として 1970 年前後からまず ILO（
「世界雇用プ
ロ グラ ム 」） で、 そして 世界 銀行 で活 発に 使わ れた 社会 会計 行列 （SAM: Social
Accounting Matrix）の研究の自然な発展であったこと、したがって SAM ベースの CGE
であったことにその最大の特色があるだろう。
本来、SAM の「社会」という語の意義は、都市−農村、富裕層−貧困層といった社
会経済的背景に沿って制度部門分類、とくに家計部門の内訳部門分類を行なうことにあ
る1。生産構造（投入産出構造）だけに基づくレオンチェフ逆行列（行列乗数）は、生
産構造と所得分配とを同時に考慮した SAM 乗数分析に拡張され、世界銀行の「分配に
配慮した成長戦略」を支える重要な道具立てとなった。そうした SAM の発展としての
CGE 分析は、SAM 乗数分析のもつ、線形性制約や固定価格制約から自由な分析を可能
にするものとして位置づけられた。その整合性を確保しながらデータを格納するフレー
ムワークとして、また、モデルを提示する手段として、そこでも SAM が最大限に利用
されたことは当然のことであった。
今回、われわれが目指す分析は、後者、すなわち、世界銀行の伝統に沿った SAM ベ
ースの CGE 分析である。ただし、アン・クルーガーの登場により、世界銀行は IMF
1
Chowdhury and Kirkpatrick[1994]
-37-
とともに悪名高い構造調整政策の実施機関になる2。われわれの目指す方向は、変質以
前の世銀 SAM 派の行なってきた分配重視の CGE である。
１．２ 環境政策の分析にとっての CGE の有効性
たとえば、環境税（炭素税）の景気への影響を分析するのなら、マクロ計量モデルが
適した分析ツールであるかもしれない。しかし、直接規制にしても間接規制にしても、
環境負荷の大幅な低減をめざす環境政策の効果を分析しようとする場合、エコノミー・
ワイドな多種の相互作用を通じて、それが相対価格や当該経済の産業構成に、さらに、
それが所得分配にまで大きな影響を与えることを考慮に入れなければならない。CGE
はそのためのひとつの最適な枠組みを与えていると考えられる。また、環境政策を分析
するための CGE には既に十分な研究蓄積がある3。
しかし、CGE 分析には、いつかの弱点も指摘されている。とくに、カリブレーショ
ンについては、改善の余地があるだろう。単年データを使って自動的になされる部分も、
そうでない部分についても、パラメーターの決定に、より実証性がほしいところである。
最近は、計量経済学的手法を取り入れる方向も試みられているが、パラメーターの変更
だけでなく、クロージャーを含むモデル設定の変更に対して、CGE 分析がもつ柔軟さ
を生かそうとする方向もありえるだろう。動学分析や金融的側面の分析が、今のところ、
未成熟であると批判されることも多いが、今回のプロジェクトにとってのダメージには、
それほど大きくないと考えられる。
１．３
分析手順
次のようなステップで分析を行った。
(1) 日本版 NAMEA(ハイブリッド勘定)データを、環境技術データを用いて、93SEEA
ヴァージョン IV.２型のデータに変換する。そうする理由については次節で議論する。
(2) 後者を世界銀行の流れを汲む CGE 分析用の SAM に変換する。帰属環境費用を含ま
ない SAM も作成する。
(3) 以上のステップで構築された SAM を使い、次のような SAM ベースの CGE を実行
する。
(3)―１）維持費用方式で計算されている仮想的費用を実際にかけるとしたら、経済の
営みはどのように変わるか。維持費用方式の帰属環境費用が波及効果を含まない、仮想
的な数値であるという批判に答えるためにも、また、環境税のような間接規制と直接規
制とを比較するという目的にとっても有益であろう。
80 年代の世界銀行の変質については、Stiglitz[2002]で活写されている。その変質の結果、
SAM 派の多くの有能な研究者が世界銀行を去って、ウォリック大学（グレアム・パイアッ
トの場合）や国際食糧政策研究機構（IFPRI、シャーマン・ロビンソンの場合）などの研究
機関に席を置くようになった。
3 Xie[1996]、Conrad[1999]などを見よ。
2
-38-
(3)―２）環境からの生産要素投入があるという仮説にたった分析を行なう。SAM を構
成する際、エコマージン（帰属環境費用）をいわば環境を「搾取」して得られた付加価
値であると考える。所得フローの面では、資本要素への支払いとまったく同様に扱う。
このような仮説のもとで、各種の比較静学分析を実行する。
(3)―３）環境税、排出権のような環境政策手段の検討を行なう。
１．４
SEEA ヴァージョン IV．２とハイブリッド勘定（日本版 NAMEA）の比較・
評価
「新しい環境・経済統合勘定について」という文書で発表された、「ハイブリッド勘
定」は、初期の、国連の暫定環境・経済統合勘定マニュアル（93SEEA）のバージョン
IV.2 に基づいた試算とは異なり、オランダ中央統計局が NAMEA
（National Accounting
Matrix including Environmental Accounts）という名前で開発した環境統計とほぼ同
じものであり、国民勘定行列（NAM）に環境負荷を記録する部分を付け加えたもので
ある。
最新の環境・経済統合勘定（ハイブリッド勘定）で取り上げられている環境負荷項目
の範囲は従来と比べて、格段に拡大しているが、むしろ、SEEA の 90 年代における試
算と最新の試算との間には、大きな枠組みのちがいがあることに注意すべきであろう。
それは、ごく単純化して言えば、次のように表現することができるであろう。
環境負荷量の測定＋経済活動との関連づけ ＝ ハイブリッド勘定
ハイブリッド勘定＋環境負荷の貨幣評価の帰属（帰属環境費用）＝
従来の（version4.2 の）環境・経済統合勘定
明らかに、統計を作成するうえで必要な情報は、従来型の環境・経済統合勘定のほう
が多い。なお、
グリーン GDP＝GDP−固定資本減耗−帰属環境費用
であるが、ハイブリッド勘定では、グリーン GDP は計算されない。また、帰属環境費
「維
用の計算にはいくつかの方式があるが、ヴァージョン IV.2 では、既に見たように、
持費用」と呼ばれる考え方に基づいた帰属環境費用の計算が行なわれる。
環境負荷の貨幣評価は、論争的な話題であり、必ずしも貨幣評価を是としない立場も
「持続可能性をもった発展」
ある。しかし、ヴァージョン IV.2 の環境・経済統合勘定は、
（sustainable development）の見地から支持しうるものと考えられる。ある企業が用
いている機械に問題があり、そのため環境を汚染するが、その機械に問題の環境汚染に
対処する装置を組み込むことにより、問題を解決（軽減）させることができるとしよう。
しかし、その企業はそうした努力を怠り、その結果として現実に環境が汚染されている
-39-
状況を考える。この場合、帰属環境費用の計算は以下のように行なわれる1。
第１図 維持費用概念の説明
持続可能性の政策目標の設定
↓
環境基準の制定（たとえば、ゼロエミッション）
｜
｜
汚染量の測定
｜
｜←汚染除去技術の想定
｜
↓
｜
｜
――→
汚染量単位あたり汚染除去費用の算定
↓
（汚染量−環境基準にもとづく汚染限界）×単位あたり除去費
用2
＝帰属環境費用
以上のような帰属環境費用の計算には、「持続可能性をもった発展」を実現するため
に環境政策当局が何をしなければならないか（どのような規制を導入しなければならな
いか）ということについて明確なイメージをもっていなければならない。その点がそう
したコミットメントのないハイブリッド勘定との大きなちがいである。環境問題の解決
のためには、そうした強いコミットメントが必要であると考えられる。
留意すべきことは、グリーン GDP は、いわば、副次的指標であり、帰属環境費用こ
そが持続的発展のための主要な指標であるという点である。帰属環境費用は、環境を犠
牲にすることによって得た付加価値（エコ付加価値）であり、それをゼロにすることが
持続的発展を実現するための政策目標となるからである。GDP とグリーン GDP の成
長率を比較することは有意義だが、GDP の成長率をグリーン GDP の成長率に置き換
えれば、後者が持続的発展の指標と考えることは誤解、指標の誤用であり、環境に大き
なダメージを与えることにつながるであろう。
最後に、関連して、国民経済計算調査会議体系整備検討委員会（2005年6月16日）に
おける河野政男委員（中央大学教授）の発言を議事録から引用しておこう3。「サテラ
イト勘定の整備ということで、環境勘定にかかわることです。これから地域版の環境勘
定について、研究を進めたいということでありますが、ぜひＮＡＭＥＡ版ばかりでなく
1
このようなエンド＝オブ＝パイプ方式の技術を想定すること以外にも、より抜本的な技術
の変更を想定することもできる。
2 汚染除去費用に非線形性が存在する可能性があるが、単純化する。なお、日本の環境・経
済統合勘定の初期の試算では、ゼロエミッションが想定されていた。
3 内閣府経済社会総合研究所[2005]。
-40-
て、ＳＥＥＡヴァージョンIV.２ですか、維持費用評価についても引き続き研究してい
ただきたいというふうに思っております。と言いますのは、実は企業会計の方で、イギ
リスの貿易産業省支援のもとに、シグマ計画という計画がありまして、そのもとでは環
境に優しい企業のいろいろな経営のためのガイドをつくっているんですが、その中にエ
ンバイロメンタル・アカウンティング・ガイドラインというのがあります。そこでは企
業が持続可能なためには、税引き後利益から維持費用を引く、企業が排出する負荷物質
について、それをゼロエミッションしたときのコスト、ゼロエミッションまでいかなく
ても、持続可能な排出量というのがあって、そこまで減らすのにかかるコストを引いた
ものが持続可能利益だということで、そういう会計決算書を示した上で単位当たりの負
荷物質の削減のための計算レートなんかをみんな載せたものがあるのです。既にこうい
うことを実施している企業も幾つかあります。私も環境省の環境会計の２００５年版の
ガイドライン、企業向けのガイドラインの作成にかかわっていましたけれども、そこで
はシグマ計画の会計計算書も参考表として載せているのですね。企業レベルでも維持費
用ということには関心を持っているところが出てきたということでありますので、ぜひ
地域版についてもヴァージョンIV.２をお忘れなくやっていただければというふうに思
っています」。
参考文献
Chowdhury, Anis and Colin Kirkpatrick[1994] Development Policy and Planning:
an introduction to models and techniques, Routledge, 1994.（嶋田晴行・瀬戸健太・
不破雅実・三重野文晴訳『発展途上国の開発政策と計画』古今書院、1997 年。）
Conrad, Klaus[1999] “Computable general equilibrium models for environmental
economics and policy analysis,” Jeroen C.J. M. van den Bergh(ed.), Handbook of
Environmental and Resource Economics, Edward Elgar.
Stiglitz, Jeseph E. [2002] Globalization and Its Discontents, 2002.（鈴木主税訳『世
）
界を不幸にしたグローバリズムの正体』、徳間書店、2002 年。
Xie, Jian[1996] Environmental Policy Analysis: A General Equilibrium Approach,
Ashgate.
内閣府経済社会総合研究所[2005]「国民経済計算調査会議第２回体系整備委員会議事
、http://www.esri.cao.go.jp/jp/sna/050616/gijiroku.pdf
録：2005 年 6 月 16 日」
-41-
２.
２．１
研究成果と今後の課題
研究成果
・維持費用概念の帰属環境費用の明確な意義づけのもとに、それを従来の日本総研推計
NAMEA データを SEEA
の範囲、方法で、
若干の加除のもとに推計することによって、
ヴァージョン 4.2 型のデータに変換した SAM を構築した。その際、CO２、SPM に
ついて、新たに帰属環境費用を推計した。さらに、その SAM に基づいた CGE モデ
ルを実行して、試験的な結果を得ることが出来た。
・NAMEA をベースとしたことにより、従来の SEEA ベースのモデル(2003 年地域学
会発表)とは異なり、モデル上の産業分類を格段にふやすことができ、モデルの内容
も充実してきた。
・帰属環境費用データを用い、環境を生産要素とみなすという Main Model では、
Closure の意味づけを明瞭にし、排出権市場とモデルとを関係づけることができたと
考えている。一方、Alternative Model では、いわゆる Greened Economy 集計量を
モデル上計測し、統計上の Green GDP との比較をすることができ、後者が対応する
前者の下限値を与えると見られることを示すことができた。
２．２
今後の課題
・データの面では、帰属環境費用の推計対象の範囲が依然狭いので、それをさらに拡大
する必要がある。
・環境負荷項目を一括して取り扱うのではなく、少なくとも CO２とその他というよう
に、環境負荷項目の細分化された取り扱いを試みる必要があるだろう。
・Main Model の経済主体の行動想定に関して、環境コンシャスネス（ペーパーにおけ
る記述を参照のこと）の想定を様々に変更してみる必要がある。
・モデルの生産構造を変更して、比較静学実験結果のロバストネスを検証してみる必要
がある。
・同モデルに排出権市場を明示的に取り入れることは興味深い。そのうえで、比較静学
実験の種類を拡大する必要がある。たとえば、排出権の海外からの購入など。
・さらに、技術のスイッチを含むモデルを工夫し、間接規制・直接規制という政策上の
課題に答えられるようなモデル構築をめざす必要があるだろう。
・Alternative Model では、投入係数変更の実験をより現実的なものにするために、環
境技術情報の収集が必要であろう。
-42-
An Environmental SAM and SAM-based CGE Modelling
for Environmental Policy Problems
Paper prepared for International Workshop for Interactive Analysis on
Economy and Environment,
Cabinet Office, the Government of Japan,
4th March 2006
Noritoshi Ariyoshi
Faculty of Environmental Studies
Nagasaki University
Itsuo Sakuma
School of Economics
Senshu University
Akihiko Taniguchi
Graduate School of Economics
Senshu University
-43-
1. Introduction
The purpose of this paper is twofold. First, concerning data aspects, it attempts to
construct a data matrix (DATA SAM) from Japanese NAMEA 1 (or hybrid accounts)
compiled by ESRI. The data matrix which the authors will formulate contains national
accounts data as well as SEEA2 (ver.IV.2) type economic-environmental data. In doing so,
imputed environmental cost, one of the main features of SEEA verIV.2, will be estimated for
each industry. It should be stressed that the estimation of imputed environmental cost or
maintenance cost is highly important because by doing so what the government should do
for the environment will be clarified and obliged at least logically. NAMEA as such, on the
contrary, dose not contain any information of this sort. This DATA SAM will be called
“DATA ENVIRONMENTAL SAM”.
Second, concerning modelling aspects, it attempts to conduct SAM-based CGE
analyses. To do this, the second version of “Environmental SAMs” must be created. This
second version of environmental SAM will be called “MODEL ENVIRONMENTAL SAM” in
which environmental and national accounts data are arranged to accommodate a specific
general equilibrium modelling structure. The authors attempt to conduct a computable
general equilibrium (CGE) analysis with this model environmental SAM as a base. Finally,
some (policy or other) simulation results will be presented and commented.
2. A general explanation of Japanese NAMEA (Hybrid Accounts)
In this section, the authors would like to give a very condensed account of Japanese
NAMEA or Hybrid Accounts recently compiled by the ESRI (Economic and Social Research
Institute), Cabinet Office of Japan. A sketch and a summary of Japanese NAMEA are given
in Figure 2.1 and Table 2.1. the simplest utline of Japanese NAMEA is as follows.
1) It is based on the original NAMEA by the Netherlands, and estimates for 90, 95,
and 00
2) It has twofold parallel structure: national accounting matrix (NAM) at monetary
term and environmental accounts (EA) at physical term.
3) It has twofold structure of EA : substance accounts and environmental problem
accounts.
Because further explanation of Japanese NAMEA is given in Ariyoshi (2006), it is
omitted in this paper.
“NAMEA” is an acronym for National Accounting Matrix including Environmental Accounts.
See de Haan, M. and S. J. Keuning [1996], Department of National Accounts[2004], and
Ariyoshi[2006].
2 “SEEA” is an acronym for System for Integrated Environmental and Economic Accounting.
See United Nations [1993].
1
-44-
-45-
B
Closing stocks
Hidden material flows
Decrease in land use
Input of natural assets
Disposal of pollutants
National accounting matrix
(on the monetary basis)
Increase in natural asets
A
Accumulation to environment
A-B
Pollutant’s contribution to environmental problems
C
Environmental problem
accounts
Closing stocks
Environmental indicators
Environmental accounts
(on the quantity basis)
Substances accounts
Increase in land uses
Opening stocks
Changes in natural assets
Opening stocks
Changes in land use
Emission of pollutants
Figure 2.1 Sketch of Japanese NAMEA
Hidden material flows
D
2
3
4
5
6
7
8
9
Production (activities)
Final consumption (purposes)
Value added (categories)
Distribution & use (sectors)
Taxes (tax categories)
Capital (sectors)
Non-financial assets (kinds)
Rest of the world
10d
10c
10n
R
CA
Hidden material flows
Reconciliation/indicators
Closing Stocks
1,095
7.4
44.2
44.2
Taxes (tax categories)
7
Capital (sectors)
82.6
11.9
36.4
National
economy
6
-98.0
134.4
2,885.8
8
Non-financial assets (kinds)
Billion Yen
Hidden material flows
in domestic area
2,829.3
-92.9
Exploitation of ROW's
natural resources
caused by imports
Processing and recycle
of pollutants
-1.0
13.0
0.0
55.3
9
2,826
5,883
0
81,241
16,813
-
-
ROW
Processing of pollutants and
exploitationof natural resources
0.0
370.0
369.8
5
Distribution & use (sectors)
Exploitation of
domestic natural
resources
Ｂ
Value added (categories)
4
Rest of the world
10b
20
45
64
-
Hidden material flows in
ROW caused by imports
0
10d 10e 10f
10g
0
-
0
-
-
10i
10j
10k
Land use
10l 10m 10n
-
0
-
-
0 72,841
0
Ａ−Ｂ
40
X1
Opening stocks
X2
16,813
Unit
Unit
Closing stocks
Environmental indicators
-1,095
40
-2,826
-5,883
-5,736
-40
0
-87,000
54,822 -81,241
-199
55,514
766
835
2,032
820
36,693
1,238,699
Changes in land use
Pressures on
environment
-40
Pressures on natural resources and
restoration of natural resources
Ａ
35,993
10h
Restoration of Natural resources
-
143 346 406
775 1,889 490 360 424,840
10c
汚染物質の排出
Emission of pollutants
221,424
1,017,275 36,609
10a
Quality
of water
Natural resources
Accumu. To env.
Global
CO2
-
11e
21,371
11f
-
11g
-
11h
-
11i
Changes in
land uses
-
-
466
-
-
17 55,514
-
-
-87,000 -5,736
-
11l
Others
40
-
40 -1,095
4,910 25,380
-40
-1,095
-40 Agricultural land
Changes in
natural
resources
Hidden material flows
-
-199 54,822
Ｄ
Changes in
land use
54,822 Changes in Forests
Chamges in
-87,000
water esources
Changes in
fishery resources -5,736
21,172
Environmental indicators
11k
4,950 25,370
11j
Contribution to regional
environmental issues
-199 Depletion of energy resources
17 Wastewate
55,514 Final disposition of waste
466 T-P & T-N
1,422 Others
1,332,945 2,242
-
-
11d
Contribution to global environmental issues
820 SO2
Making indicators
Ｃ
-
11c
Changes in natural resources
The potential contributions to Global warming, acidification and
Eutrophication are expressed in GWP, AEQ and EEQ, respectively.
Global Warming Potencials (GWP) CO2 ： 1, N2O：310, CH4：21
Acid Equivalents (AEQ) Nox ： 0.7, So2：１, NH3
Eutrophication Equivalents (EEQ) P ： 3.06, N：0.42
-
11b
94,246 Others
1,238,699
-
11a
Regional
Environmental problems
Note) This is a summary table of the attachment#1-3 of New System of Integrated Environmental and Economic Accounting, Department of National Accounts, Economic and Social Research Institute (ESRI), Cabinet Office of
Japan, 2004 (http://www.esri.cao.go.jp
Unit
10m
10l
5,736
18,019
199
405,319
0
0
38.7
468.0
10k
Land Agricultural land
use Others (5)
3
Final consumption (purpose)
434.9 369.8
87,000
10i
Forest (volume)
47.9
5.1
-0.2
941.5
2
Goods & services (products)
10j
10h
Energy(3)
Water resources
(wateruse)
Fishery resources
(marine products)
10g
Waste(2)
Quality T-P＆T-N 10e
of water Wastewater 10f
Acidifi- SO2
cation Others(2)
10b
10a
1
Goods & services (products)
Green- CO2
house Others(5)
OA
Production (activities)
1
Air
Greenhouse effects Acidification
Pollutants
Substances
Accumulation of substances
to the environment
: Cells for environmental indicator
Changes of natural resources
in ROW caused by import
: Cells for environmental flow data
SO2
Account
(classification)
Air
Opening stocks
Pollutions
Substances
Natural resources
Wastewater
1,000 ｔ
CO2
1,000
tonnes
-CO2
T-P＆T-N
1,000 ｔ
Others (5)
1,000 t
-N2O
Others (2)
1,000 t
-NOX
1,000t
-SO2
Energy (3)
PJ
Acidification
1,000t
-SO2
1,000 ｔ
Material flows
3-
Monetary flows
-PO4 .
1,000 ｔ
Waste (2)
1,000 ｔ
Eutrophication
Table 2.1 Japanese NAMEA (2000) - Summary Table -
3-
Forests (volume)
1,000ｍ3
Waste
1,000
ｔonnes
Water (water use)
1 million
ｍ3
Energy resources
PJ
Fishery resources
(marine products)
1,000 ｔ
Forests
1,000
ｍ3
Agricultural lamd
1,000 ha
Water
1 million
ｍ3
Othrs (5)
1,000 ha
Land for housing
1,000
ha
Hidden material flows
1 million
ｔonnes
Wastewater
-PO4 .
Other lands
1,000
ha
Greenhouse effects
1,000
tonnes
-CO2
Fish
1,000
ｔonnes
Hidden material flows
1 million
ｔonnes
-46-
3. From NAMEA to a SEEA-type SAM with Monetary Valuation of Environmental Pressure
Table 3.1 is a summary version of a SEEA-type SAM with monetary valuation of
environmental pressure. In order to obtain this table, the two processes are needed. To
begin with, imputed environmental costs are calculated by maintenance cost methods using
the estimates of environmental pressure in Japanese NAMEA for 1995 and the additional
pollutants (SPM). Then, table 3.1 is obtained by incorporating these imputed
environmental costs into Japanese NAMEA.
Table 3.2 shows imputed environmental cost (eco-margins) by origin of industries
calculated by the maintenance cost method. According to the table, the largest eco-margin
is found in non-metallic mineral products.
Table 3.3 shows imputed environmental cost by type of environmental pressure.
According to the table, CO2 has the largest eco-margin.
The most notable difference between the two statistical efforts made by the
Government of Japan (new NAMEA and old SEEA) might be that in their positions on
monetary valuation of environmental pressure. However, it is no use listing up pros and
cons on monetary valuation of environmental pressure in general. Thus, the authors are
pro the maintenance cost concept, but con the concept of green GDP. If green GDP were
used as a substitute for GDP in the ordinary sense, it might justify inappropriate
growth-accelerating policies which may worsen the environment.
In this section, the authors will attempt to construct a SEEA-ver.IV.2-type matrix
from Japanese NAMEA. In doing so, imputed environmental cost must be estimated by
using the maintenance cost concept.
This imputation is considered to be highly important because it can clarify what the
government should do for the environment as noted before. In this paper, however, only
imputed environmental cost the origin of which is industries (=market producers in 93SNA)
are considered, because it is necessary to take the extension of production boundary into
consideration in order to deal with consumption-related imputed environmental costs.3
Except for CO2 , the assumption of “zero-emission” is posited in order to calculate
imputed environmental cost about each environmental pressure category. Concerning CO2,
the 94% of the emission level of CO2 in 1990 is assumed, according to the Kyoto Protocol.
Two tables below summarise imputed environmental costs (=eco-margins) by sources and
by environmental pressure types (matters). More detailed tables including a table for
description of the method of calculation are attached at the end of the paper.
“Shifting” was a device in 93SEEA to deal with imputed environmental cost originated in
non-producers.
3
-47-
Ａｃｃｏｕｎｔ
（Classification）
C a p ita l a c c o u n ts
In c o m e
a c c o u n ts
Rest of the world
1
4
3
14 38272.4
4051.6
973.1
369345.0
4
5
5901.3
6
45313.1
9
15880.6
10171.8
86965.8
-86965.8
171.8
-11428.0
11428.0
-6969.8
-88912.7
134225.8
8
6969.8
-204.2
6105.5
7
-68568.0
(-) sum of imputed environmental costs
valuated by maintenance cost approach
57549.1
349633.2
Imputed environmental cost
valuated by maintenance cost
approach
349633.1
3
Non-produced
assets
(types)
68568.0
33372.6
458525.0
431040.2
2
Use and
Goods ＆ Productio Final
Generation
distribution accumulatio Environment
services n(activitie consumptio of income
of income n(institution al protection Other
(products)
s)
n(purpose) (items of
produced
(institutional al sectors) facilities
value added)
assets
sector)
(types)
Capital accounts
produced assets
10
11
12
13
E c o - m a rg in
(-) sum of deterioration of
non-produced assets by
production activities
Deterioration of non-produced
assets by production activities
A ir
p o llu tio n s
Imputed environmental costs
Environmental impacts
Q u a lity o f
w a te r
Income accounts
W a s te
Table 3.1 a SEEA-type SAM with Monetary Valuation of Environmental Pressures (1995)
2 922938.0
1
Use and distribution of income
(institutional sector)
5
accumulation
(institutional sectors)
6
Environmental protection
facilities(types)
7
Other produced assets
(types)
8
Non-produced assets(types)
9
Discharge of residuals
(CO2,SOX,NOX,SPM
10
etc)
Quality of water
(T-P,T-N,waste water) 11
Waste
12
(final disposition)
Eco-margin((-)sum of imputed
environmental costs)
13
Generation of income
(items of value added)
Final consumption(purpose)
Production(activities)
Goods ＆ services(products)
p ro d u c e d
a s s e ts
Im p u te d e n v iro n m e n ta l
c o s ts
E n v ir o n m e n ta l
im p a c ts
-48-
-214.4
19372.2
108.7
45230.1
14
Rest of
the
world
Table 3.2 Imputed Environmental Cost by origin of industries
Distribution
Eco-margins
of
Industries
Eco-margin
(Billion yen)
(%)
Distribution
of Value
Added (%)
2295.4
2.6
2.0
91.8
0.1
0.2
940.4
1.1
2.8
1947.8
2.2
0.3
237.0
0.3
0.7
1521.7
1.7
2.1
255.9
0.3
1.2
Non-metallic mineral products
29863.3
34.3
1.0
Iron and steel
11573.6
13.3
1.3
443.7
0.5
0.4
Fabricated metal products and Machinery
2260.8
2.6
10.7
Other Manufacturing
5560.7
6.4
4.1
Construction
3531.9
4.1
8.7
Electricity, gas and water supply
8035.8
9.2
2.9
Wholesale and retail trade
2829.0
3.3
16.3
Finance and insurance
1150.7
1.3
6.3
Real estate
2229.8
2.6
12.8
Transport and communications
8786.8
10.1
7.6
Service activities
3409.6
3.9
18.8
Total Industries
86965.8
100.0
100.2
Agriculture ,forestry and fishing
Mining
Food products and beverages
Textiles
Pulp, paper and paper products
Chemicals
Petroleum and coal products
Non-ferrous metals
-49-
Table 3.3 Imputed Environmental Cost by type of environmental pressure
Types of Environmental Pressure
Eco-margins
CO2
61033.4
N2O
11.7
CH4
4.5
HFCs
88.2
PFCs
50.7
SF6
73.7
NOX
4984.7
SO2
2261.6
SPM
59.4
Green House Effects
Discharges
Acidification
Quality of Water
Wastes
T-P
3390.2
T-N
2552.2
Waste Water
1027.4
11428.0
Final Disposition
Total
86965.8
4. A CGE Analysis
The authors will attempt to conduct a CGE analysis in this section. A few words may be
necessary about the terminology before going into the subject.
4-1.AGE and CGE
Although the term CGE (Computable General Equilibrium) is often used as a synonym
for AGE (Applied General Equilibrium), these two terms are considered to be reflecting
different genealogy of a sort. The latter term suggests that the study in question lies in the
tradition of Scarf and his successors. The former term would identify the study as being in
the World Bank’s tradition of development modelling. As is well known, Scarf and his
colleagues attempted to find the exact location of a general equilibrium not just its
existence. They developed and improved algorithms for finding a fixed point so that they
could conduct various policy simulations.
On the other hand, the modellers in the World Bank who had used SAM multiplier
analyses to draw up development policies felt that their tool should be more flexible. In
fact, the SAM multiplier analyses confine their application only in the cases where the
prices are constant and production functions and consumption functions are linear. The
computation of general equilibrium has made it possible for them to generalise SAM
-50-
multiplier analyses to SAM-based CGE analyses.
The CGE models in this tradition have several common features. Of course, one
common formal feature is that they are SAM-based. SAMs (Social Accounting Matrices)
are in the centre of World Bank’s modelling attempts. As is well known, Professor
Graham Pyatt played a leading role in development of SAM-based CGE modelling, in
which much use was made of SAMs both as a framework of data housing and model
presentation, whence consistency of modelling as well as data set was assured.4
It is often said that S (social) in “SAM” suggests something “social” is involved in the
data set. For example, the sub-sectoring of the “households” may be done so that the
incomes of the sub-sectors of household sector may be analysed and forecast. In fact, the
development strategy of the World Bank until the middle of 80’s was characterised by its
emphasis on income distribution. Thus, one more common feature of the CGE models,
substantive rather than formal, may their focus on income distribution. In our modelling,
however, this feature (detailed institutional distribution rather than factorial
distribution) does not appear so explicitly because necessary sub-sectoring is not done in
our modelling due to lack of data. A seemingly closely related feature may be that they do
not often rely upon utility-function-based welfare measures to evaluate alternative
economic situations which alternative policy schemes would lead to. Instead, distribution
measures like the Gini coefficient or others are often used.
Among other features is a distinctive production structure used in the models, in
which wide use is made of the so-called Armington assumption and CET (Constant
Elasticity of Transformation) production functions. See Figure4.1 below.
4-2. Features of the modelling in the present paper
The main purpose of estimating imputed environmental cost is to make clear what
the environmental policy target should be. The policy makers should use whatever
regulation measures are necessary to make the eco-margin measure reduce to zero. The
policy measures should be those designed to reduce the emission by the amount that is
indicated through the measurement of the imputed environmental cost. The policy
measures may be direct policy instruments or indirect ones to reduce the emission. Thus,
naturally, the environmental SAM modelling is strongly policy-oriented.
The authors attempt to construct two alternative models. They are both SAM-based
CGE models for environmental policy analyses.
In the main model, certain policy measures are incorporated into the model from the
very first. They could be interpreted as emission permits of a sort that should be paid for
by the producers who want to use the environment. The environment may be treated as
4
See Pyatt[1988].
-51-
one of the three production factors the remaining two of which are capital and labour as
usual.5 The environment may be considered to function as a sink of a sort. The authors
introduce a “fixed price” closure rule for the production factor instead of “fixed quantity”
closure rules for the other two production factors.
There is a thought behind this treatment. The domestic equilibrium price of the
permit will be determined at the level of imputed environmental cost, estimated
according to the concept of “maintenance cost.” Because the firms will not pay for the
permit if they can reduce the emission by spending actual environmental protection cost
which equals to maintenance cost and is less than the value of the permit.
The alternative model may be called “Ordinary Economy” model because the
environment does not play any role of a production factor in this model. In other words,
the economy is an ordinary (or actual) one without any environmental imputation. By
using this model, the consequence of a certain regulation measure which is designed to
reduce the emission by the amount designated by the policy target can be projected. For
example, the consequence of direct regulation of the production processes of the firms
discharging residuals in an inappropriate manner might be analysed.
In the main models, the payment flow (for the environment as a factor of production)
is received as if it is owned by the capital owners, who are characterised as those who
take any surplus gain in the production process. The government in turn takes up the
sum as a kind of transfer payment. But in a base equilibrium, it will be transferred back
to the owners of capital to reproduce the actual data in 1995.
This needs some excuse actually. Thus, imputed environmental cost is the part of
value added of the producers which they get through exploiting the environment, so to
speak. Firms doing so may not be so aware of the environmental pressure they have
made. At the same time, however, it is also the case that the present state of the economy
is not the result of “zero environmental awareness.” In fact, there is, at present, so much
pressure on the firms that are responsible to any ill-usage of the environment in the
society. For example, the possibility of damage payment in the future may cause the
firms to act in a more deliberate manner. Any new system for environmental protection
will force the firms to adjust to it. Thus, the models assume that the firms are just ready
for the introduction of a national tradable permit system although in reality they do not
make any net payment about it.
But, this hypothesis may be considered to assume too “environment-conscious” firms
Rather than treating the environment as a production factor, a possible alternative model
that we might suggest (but not fully developed here) will be the model where the emission of
residuals are treated as a joint production of a sort which might be formulated in terms of
“CET” technologies. This treatment seems to fit quite well with the logical framework of
NAMEA.
5
-52-
in the model. It may be better to think that the present size of eco-margin may be the
result of the behaviour of the firms which respond to not full but a half or even one third
of the domestic equilibrium price of tradable permits. Thus, it might be more plausible to
assume less “environment-conscious” behaviour of firms in the model. In any case, it will
be easy to adapt or reinterpret the model in the way suggested here. However, it should
be careful about the implementation and interpretation of the simulation results
Production part of the model, which is one of typical World-Bank-type models except
for the treatment of the environment, may be schematised in Fig. 4.1.
A few comments should be in order. (1) The environmental production factor is
combined with mixed production factors by using CES production function form. Aside
from a parameter automatically calibrated, the modeller(s) need(s) to set one parameter
of the CES production function, namely, the elasticity of substitution. The authors set it
to be 1.5.1 (2) Domestic goods (and services) stage 1 are domestically produced goods
some of which may be further transformed to exports, the rest being (transformed into)
domestic goods (and services) stage 2. The CET transformation function indicates the
process. (3) Domestic goods (and services) stage 2 combined with imports are
transformed into composite goods (and services) that are to be used for the final uses and
intermediate consumption.
A sketch of the model SAM as a whole is illustrated below in Table 4.1
6
If the elasticity parameter is set otherwise, the results will change, but only slightly.
- 53 -
Table 4.1 Model SAM sketched
Factors
Institutions
Savings.-
including
Investment.
Com.
Activities.
ROW
Factor
Factor
Payment to
Payment
Capital,
From the
Labour,, and
ROW
Company
a/c
Factors
Environment
Institutions
Primary
Transfer
Transfer
including
Distribution
payments
from
Company
Income Generation
between
ROW
account
on the Assumption
Institutions
that payment to
including
the environment
those related
goes as if it is
with
owned by the
emission
capital owners
permits
Savings-
Savings
Investment
Commodities
Final
Capital
Intermediate
Consumption
Formation
Input (U)
Activities
Output
(V)
Rest of the
Factor payment to
Transfer to
Current
World
the ROW
ROW
Balance
-54-
Imports
Exports
4-3. Results for the main model
The increase in the endowments of capital or labour leads to the decrease in
imputed environmental cost (eco-margin). In one sense, the environment is ill-used just
to save other factors. But size of pressure of the 10% increase in each factor endowment
to the industries is different industry by industry. Concerning labour, a large decrease in
eco-margin is found in construction, transport and communication, service, etc.
Concerning 10% increase in capital, a large decrease in eco-margin is found in real estate,
agriculture, wholesale and retail trade, etc.
The 10% increase in labour endowment has a larger effect on the reduction of
eco-margin than the same proportional increase in capital endowment.
The result of the experiment about 10% increase in the price of environment as a
production factor may not seem to be so definite. Eco-margins decrease in some
industries but increase in other industries.
The results of selected experiments will be summarised as follows
Table 4.2 Selected Results of the Main Model
% changes in
% changes in
% changes in
eco-margin due to
eco-margin due
10% increase in
to 10% increase
Labour
in Capital
Agriculture, forestry and fishing
-5.4
-7.9
1.3
Manufacturing
-5.4
-2.9
3.9
-13.5
-3.3
-0.3
Electricity, gas and water supply
-6.2
0.6
0.0
Wholesale and retail trade
-9.2
-5.5
-1.5
Finance and insurance
-7.7
-4.2
-1.5
Real estate
-3.5
-8.4
-1.2
Transport and communications
-10.7
-1.0
1.2
Service activities
-10.5
-5.1
-0.5
Industries
Construction
-55-
eco-margin due to
10% increase in
environmental factor
price
4-4. Results for the ordinary economy model
In addition to the main model, an alternative model is formulated to analyse an
ordinary or actual economy without environmental imputation. Policy makers may be
inclined to think that it is necessary to introduce direct rather than indirect regulation
measures in order to reduce the imputed environmental cost known quite immediately to
zero. In other words, the imputed cost in the main model needs to be borne actually by
the producers of the alternative model. Thus, they must spend some additional cost for
the environmental protection instead of discharging residuals.
On this assumption, the working of the economy will be simulated by using this
model. Additionally, so-called “Greened GDP” or more adequately “Greened Economy
GDP” in the terminology of SEEA2003 draft manual will be calculated although “greened
economy SAM” may be more meaningful.
According to the authors’ model, Greened Economy GDP seems to be greater than
Green GDP by the amount depending on the assumptions made.
Some words may be necessary here to compare the two concepts, “Green GDP”
(GDP minus Fixed Capital Consumption minus Imputed Environmental Cost) and
“Greened Economy GDP.” One typical criticism on the former concept may be that it does
not take into account various percussion and repercussion of employing one or other
residual reduction measures. Some authors advocated “Greened Economy GDP” instead
of Green GDP to respond to the criticism. Numerical comparisons between the two are
found in the table below.
Table4.3 Green GDP and Greened Economy GDP
（Billion yen）
GDP AT FACTOR COST
469,870.3
GDP AT MARKET PRICES
509,969.1
FIXED CAPITAL CONSUMPTION
89,116.9
NDP AT MARKET PRICES
420,852.2
ECO‐MARGIN in the model
86,025.0
GREEN GDP
334,827.2
GREENED ECONOMY GDP (CASE 1)
407,862.4
at constant prices
GREENED ECONOMY GDP (CASE 2)
392,774.6
at constant prices
GREENED ECONOMY GDP (CASE 3)
400,739.7
at constant prices
CASE 1: Imputed environmental cost replaced by INTERMEDIATE INPUT
CASE 2: Imputed environmental cost replaced by FACTOR INPUT
CASE 3: Imputed environmental cost replaced by Mixed INTERMEDIATE and FACTOR
INPUT
-56-
5. Closing Remarks
Firstly, a controversial problem over “direct” vs. “indirect” regulations will be
spotlighted by the modeling of this sort. Both types of regulations use “market
mechanisms.” The latter induces the economic agents to change their behaviors by using
the mechanisms to achieve the policy targets. The former regulates discharging of
residuals directly and the market mechanisms work afterwards to achieve a new
equilibrium position in which the policy targets are met.
The modelling seems to be insufficient to get to any decisive conclusion about the
problem raised. Because the key to the problem may be whether any switch of techniques
as well as shift in industrial structure is triggered by the policy change. The main model
excludes any switch of techniques from consideration. However, plausibly, indirect
regulations may bring about insufficient, if not nil, switch of techniques for the
accomplishment of the policy target. If this is the case, direct regulation may be better
way to achieve the target, for additional consequences brought about by indirect
regulations may not necessarily be desirable from the viewpoint of income distribution.
Secondly, needless to say, more disaggregated treatment for the environmental
pressure categories should be necessary. Green house gases such as CO2 and other
residuals cause different environmental problems that need to be solved differently.
Lastly, it should be stressed again that monetary valuation of environmental
pressure according to the concept of maintenance cost is important in that it not merely
gives an estimate of the equilibrium price of an emission permit but clarifies the
responsibility of the government for the environment. For, the estimation of imputed
environmental costs clearly has an implication that they need to be reduced to zero.
References
Ariyoshi, N. (2006), "The Development of Japanese NAMEA" the paper prepared for
International Workshop for Interactive Analysis on Economy and Environment,
Cabinet Office, the Government of Japan, 4th March 2006.
De Haan, M. and S. J. Keuning (1996), "Taking the Environment into Account：The
NAMEA Approach," The Review of Income and Wealth, Series 42, No.2,
pp.131-148.
Department of National Accounts, Economic and Social Research Institute (ESRI),
Cabinet Office of Japan (2004), “New System of Integrated Environment and
Economic Accounting:
Trial Calculation on
-57-
Hybrid Accounting System
Integrating
Environmental
Pressures
and
Economic
Activities,
http://www.esri.cao.go.jp/en/sna/menu.html.
.
Pyatt, G. (1985), " A SAM Approach to Modeling,” Journal of Policy Modeling, vol.10, pp.
327-352.
United Nations (1993), Handbook of National Accounting: Integrated Environmental and
Economic Accounting, Interim version, United Nations (ST/ESA/STAT/SER.F/61).
-58-
-59-
Unit cost of
CO2
ohter
environmental Unit costs for
pressures
maintenance
CO2
Unit costs for maintenance
(total emission less permissible level of emission)
×
maintenance
costs
depreciation
0.1
ratio of salvage value
maintenance costs data known for the environmental
facilities for the mitigation of air polltion /capital
formation in the environ,mental facilities for the
mitigation of air polltuion
ratio of maintenance
costs to investment
processing capacity of each type of environmental ficilities×the removal rate
The amount known concerning the investment in
environmental facilities for the mitigation of air
plollution ×accumulative gross output of the specific
vategory of environmental facilities／accumulative gross
output of environmental facilitie for the mitigation of
air polltion
plant and equipment
investment for a
specific category of
environmental
facilities
plant and equipment investment of a specific category of environmental
facilities×ratio of maintenance costs to investment
accumulative plant and equipment investment known for the
mitigation of air pollution × accumulative gross output
of the specific category of environmental facilities
/accumulative gross output of environmental facilities for
the mitigation of air pollution
accumulative plant and
equipment investment
for specific category
of environmental
facilities）
accumulative plant and equipment investment for a specific category of
environmental facilities）×（１−ratio of salvage value）／service
life(7 years)
depreciation + maintenance costs
the relevant costs incurred about the elimination of environmental pressures totalled for all gotegories of environmental
facilities /environmental pressures eliminated totalled for all categories of environmental facilities
environmental pressure eliminated stated in the environmental reports
maintenance cost stated in the environmental reports
estimated by PIM method using investment data stated in the environmental reports of selected companies with service life being 7 years
（Depreciation＋maintenance costs）/(CO2 emission eliminated)
＝
Note1:The Kyoto Protocol set the permissible emmission level of CO2, while zero emmission is assumed about the other environmental pressure items..
Note2: The estimation method concerning environmental pressures other than CO2, the same method as used in the ESRI's compilation of SEEA accounts is used...
Environmental
pressures
eliminated
elimination cost
Eliminated CO2
environmental
pressures
Maintenance costs
Depreciatin
Inputed
environmental costs
Appendix 1 Estimation Method of Imputed environmental costs: Outline
-60-
Eco-Margin
Wastes
Water Use
Discharges
6.25
4.70
1.89
30.95
91.81
67.81
51.05
20.55
301.37
2295.40
T-N
Waste Water
Final Disposituon
T-P
0.07
7.09
0.62
23.00
NOX
SPM
0.14
1.41
SF6
0.04
0.10
0.97
PFCs
5.18
0.17
1.68
0.22
N2O
HFCs
0.02
1815.00
CO2
0.01
46.83
1
0.09
2
A griculture ,forestry and fishing
CH4
M ining
SO2
Acidification
Green House
Effect
Industries
Food products and beverages
940.43
324.39
28.37
70.47
93.61
0.67
2.06
2.07
2.96
2.03
3.54
0.18
0.47
409.59
3
T extiles
Pulp ,paper and paper products
92.89
7.47
18.57
24.66
2.05
2.15
4.04
0.82
0.56
0.98
0.05
0.13
82.59
5
258.98
21.50
53.41
70.95
1.75
3.07
7.25
2.26
1.56
2.71
0.14
0.36
1097.72
6
C hem icals
1947.83 236.95 1521.67
53.08
3.42
8.49
11.27
0.36
0.72
1.24
0.34
0.23
0.41
0.02
0.05
1868.20
4
Petroleum and coal products
255.89
114.31
11.84
29.42
39.08
0.81
1.98
5.59
0.91
0.63
1.09
0.06
0.15
50.03
7
N on-m etallic m ineral products
9
Iron and steel
10
188.09
12.95
32.17
42.74
2.09
2.74
9.22
1.70
1.17
2.04
0.10
0.27
73.15
5.04
12.51
16.62
0.27
0.63
1.35
0.66
0.46
0.79
0.04
0.11
Fabricated m etal products and M achinery
1466.84
110.15
273.64
363.49
0.70
0.76
3.63
12.09
8.31
14.47
0.74
1.92
4.07
11
O ther M anufacturing
C onstruction
13
14
E lectricity ,gas and w ater supply
546.16
41.67
103.51
137.49
0.37
0.67
1.35
4.05
2.79
4.85
0.25
0.64
1198.33
89.80
223.08
296.33
2.58
0.55
14.40
7.62
5.24
9.11
0.46
1.21
310.10
29.31
72.81
96.71
9.10
10.32
29.22
1.98
1.36
2.37
0.12
0.31
4716.92 1683.17 7472.13
12
W holesale and retail trade
1610.20
166.64
413.96
549.89
0.29
1.39
1.35
9.23
6.35
11.05
0.56
1.47
56.66
15
Finance and insurance
685.09
64.42
160.03
212.58
0.02
0.10
3.44
2.36
4.11
0.21
0.55
17.77
16
1291.09
131.34
326.27
433.41
0.11
0.13
0.52
5.63
3.87
6.74
0.34
0.90
29.48
17
R eal estate
29863.31 11573.56 443.69 2260.81 5560.71 3531.88 8035.82 2829.04 1150.69 2229.82
120.91
9.72
24.14
32.07
1.82
1.39
17.92
0.84
0.57
1.00
0.05
0.13
29652.74 11278.28 332.08
8
N on-ferrous m etals
Appendix 2 Eco-margin
T ransport and com m unications
8786.80
803.12
77.54
192.62
255.86
28.36
2225.94
4856.99
4.66
3.20
5.58
0.28
0.74
331.91
18
Service activities
3409.65
1958.90
193.77
481.37
639.43
0.92
1.94
4.87
12.95
8.90
15.49
0.79
2.06
88.27
19
86965.75
11427.96
1027.39
2552.22
3390.24
59.42
2261.64
4984.73
73.70
50.65
88.18
4.49
11.71
61033.42
Total by tyoe of residuals