Low Carbon Development Strategy For Mexico - An Input-Output Analysis PDF

Low Carbon Development Strategy for
Mxico: An Input-Output Analysis

Final report

Low Carbon Development Strategy for Mxico: An Input-
Output Analysis
Final report

Responsable: Dr. Pablo Ruiz Npoles

Este reporte fue elaborado por el Dr. Pablo Ruiz Npoles. El enfoque, la metodologa, las opiniones
y conclusiones son responsabilidad exclusiva del autor.
Este informe ha sido posible gracias al apoyo del Programa de las Naciones Unidas para el Medio
Ambiente (PNUMA) y la Agencia Francesa para el Desarrollo (AFD). Su contenido es
responsabilidad del autor y no refleja necesariamente el punto de vista del PNUMA, de la AFD, de
la Secretara de Medio Ambiente y Recursos Naturales (SEMARNAT) o del Gobierno de los
Estados Unidos Mexicanos.

LOW CARBON DEVELOPMENT STRATEGY FOR
MEXICO: AN INPUT-OUTPUT ANALYSIS

Final Report

Pablo Ruiz Npoles

February, 2012

Low Carbon Strategy for Mexico: I-O Model. Final Report

ii

This report was prepared by Dr Pablo Ruiz Npoles, Faculty of Economics, National
Autonomous University of Mexico (UNAM) at the request of the Mexican Ministry of
Environment and Natural Resources (SEMARNAT). Technical support was provided by
the United Nations Environment Programme (UNEP) and the French Agency for
Development (AFD).

The finding, interpretations, and conclusions expressed in this report are entirely those of
the author and should not be attributed in any manner to UNEP and AFD. UNEP and AFD
do not guarantee the accuracy of the data included in this publication and accepts no
responsibility whatsoever for any consequences of their use.

This publication may be reproduced in whole or in part and in any form for educational or
non-profit purposes without special permission from the copyright holder, provided
acknowledgement of the source is made. UNEP would appreciate receiving a copy of any
publication that uses this publication as a source.

No use of this publication may be made for resale or for any other commercial purpose
whatsoever without prior permission in writing from the United Nations Environment
Programme.

Financial support was generously provided by the government of Norway and France.


iii

Acknowledgements

The author of this study wishes to express his gratitude to his colleagues Martn Puchet and
Valentn Sols, for their modelling advice, to Rosa Gmez, Javier Castaeda, Eduardo
Moreno and Fernando Pineda, students at UNAM, for their technical assistantship. Special
thanks to Terry Barker, Annela Anger, Douglas Crowford and Serban Scrieciu members of
the Cambridge Centre for Climate Change Mitigation Research (4CMR) for their valuable
comments, thanks also to Carl Bernadac and an anonymous reviewer, both from the French
Agency for Development (AFD) and at last but not least to Noriko Yamada from the
United Nations Environment Programme (UNEP).


iv

CONTENTS

I. Introduction ........................................................................................................................ 1

II. Review of selected literature ............................................................................................. 4

1. Climate change, causes and mitigation policies ............................................................. 4
2. Policy instruments and economic models ...................................................................... 5
3. General Equilibrium models ........................................................................................... 6
4. Econometric models ..................................................................................................... 15
5. Applied Environmental Input-Output models .............................................................. 16

III. A Comparison Mexico-Canada ..................................................................................... 22

1. Method for estimating comparison indicators .............................................................. 22
2. Estimation results ......................................................................................................... 23

IV. GHG emissions and strategic sectors in the Mexican economy ................................... 27

1. Strategic or Key economic sectors ............................................................................... 27
2. Main IOM sectors emitting GHG ................................................................................. 34
3. Key and High Pollutant sectors .................................................................................... 36

V. Environment Input-Output Model for Mexico ............................................................... 38

1. Environment Input-Output models ............................................................................... 38
2. Pollution abatement and Technological change models............................................... 39
3. Objectives of the Mexican EIO Model ......................................................................... 41
4. The model and the scenarios ........................................................................................ 42

VI. Technology changes and GHG emissions reduction: the model results ....................... 46

1. Base line or Business as Usual 2008 to 2020 ........................................................... 46
2. Introduction of technological changes and reductions of GHG emissions ........................ 48
3. Changes in GHG emissions by sector ................................................................................ 50
4. Technical change and output growth .................................................................................. 52
5. The Costs of Pollution model ............................................................................................. 53

VII. Summary, Conclusions and Policy recommendations ................................................. 54

1. Summary and Conclusions ........................................................................................... 54
2. Lessons from the policy debates ................................................................................... 59
3. Policy recommendations derived from this study ........................................................ 62
4. Further research ............................................................................................................ 63

Bibliography ........................................................................................................................ 64

v

LIST OF TABLES AND GRAPHS

Tables

Table 1 Canada GGH Emissions and GHG reduction expenses ......................................... 25

Table 2 Mexico GHG Emissions and Estimated Reduction Costs ...................................... 26

Table 3 North American Industrial Classification System, Input-Output Industry
Classification ....................................................................................................................... 28

Table 4 23 Selected Industries of the 2008 I-O Matrix of Mexico, Backward and
Forward Linkages Rasmussen Method................................................................................ 31

Table 5 23 Selected Industries of the 2008 I-O Matrix of Mexico, Backward and
Forward Linkages Impacts by Extraction Method .............................................................. 33

Table 6 GHG Emissions by Sector in Mexico 2008, Selected Industries ........................... 35

Table 7 Main GHG Emissions Coefficients by Sector in Mexico 2008 ............................. 36

Table 8 Mexico Estimated GHG Emissions Trajectories .................................................... 37

Table 9 Mexico Estimated GHG Emissions with Technical Change .................................. 49

Table 10 Estimates of GHG emissions variations 2008-2020 by sector in
Selected Industries ............................................................................................................... 51

Table 11 Estimated Gross Output 2008-2020 ..................................................................... 52

Table 12 Estimated Gross Output and Pollution Costs for 2015 ......................................... 55

Graphs

Graph 1 Mexico: GHG Emissions BAU 2008-2020 ........................................................... 47

Graph 2 Mexico: GHG Emissions without and with technological change, 2008-2020 .... 50

Graph 3 Mexico: Gross Output without and with technological change, 2008-2020 ......... 53


1

I. INTRODUCTION

Man-made or anthropogenic climate change is defined as: a change of climate which is
attributed directly or indirectly to human activity that alters the composition of the global
atmosphere and which is in addition to natural climate variability observed over
comparable time periods (IPCC, 2007, Annex II, p.78). It is in part the result of the
atmospheric concentrations of greenhouse gases (GHG). They are those gaseous
constituents of the atmosphere, both natural and anthropogenic, that absorb and emit
radiation at specific wavelengths within the spectrum of thermal infrared radiation emitted
by the Earths surface, the atmosphere itself, and by clouds. This property causes the
greenhouse effect. Water vapour (H
2
O), carbon dioxide (CO
2
), nitrous oxide (N
2
O),
methane (CH
4
) and ozone (O
3
) are the primary greenhouse gases in the Earths atmosphere.
Besides CO
2
, N
2
O and CH
4
, the Kyoto Protocol deals with the GHG sulphur hexafluoride
(SF
6
), hydro-fluorocarbons (HFCs) and per-fluorocarbons (PFCs), (IPCC, 2007, Annex II,
p.82). GHG are primarily produced by the combustion of fossil fuels, agriculture, land-use
changes and production of materials such as cement, as well as the burning of waste.
Climate Change consists of a gradual increase in the planets temperature, rise in
sea levels and changes in its rainfall patterns, as well as in the frequency, magnitude and
intensity of extreme weather events such as droughts and floods. Although this tendency
has been scientifically verified, there is still some degree of uncertainty about the
magnitude and velocity of these changes at a regional scale. However, based on the current
state of knowledge it is possible to identify some of the cause-effect chain relations
between GHG sources, GHG emissions, global warming and its climatic consequences.
This allows us to foresee various future scenarios for the economy, based on which
we can assess, from an economic perspective, the possible consequences of climate change
and the alternative options for adaptation and mitigation policies, in order to face the
problem. Mitigation has been defined as: the technological change and substitution that
reduce resource inputs and emissions per unit of output. Although several social, economic
and technological policies would produce an emission reduction, with respect to Climate
Change, mitigation means implementing policies to reduce greenhouse gas emissions and
enhance sinks (IPCC, 2007, Annex II, p.84). However, as some expert has pointed out it is

2

not only emissions intensity reduction (i.e., GHG emissions per unit of output) but also
absolute emissions reduction which is important in mitigation.
In general, mitigation policies aim to the reduction of fossils fuels consumption and
substitution, towards low-carbon sources (and/or the capture and storage of carbon from
emissions) therefore the factors that cause it must be dealt with. These factors are mainly:
population dynamics, urbanization, production and consumption increases; energy
efficiency and technology innovation tendencies, as well as the economic structure, in each
country. All these factors are, one way or another, related to economic activity in a broad
sense: production, trading, consumption and investment.
From an economic perspective, in order to design a mitigation scenario, it is
necessary to identify those economic sectors of production, or industries, which directly or
indirectly generate GHG emissions becoming, therefore, the sectors that call for special
attention; these are key sectors for mitigation. This can be seen as a supply-side view,
though, since there is also a demand-side of the problem which is related to consumption,
investment and exporting and could also be subject to mitigation policy actions.
In turn, the costs of mitigation measures depend on various local circumstances, for
example, in the case of production, the specific form of economic growth and the
introduction of technology developments in the production process aimed to reduce GHG
emissions. Besides, climate change mitigation impacts are unevenly distributed among
sectors and depend on the direct or indirect use of fossil fuels combustion of each and every
sector of the economy. In short, the economic costs of climate change mitigation depend
fundamentally on both, the energy-use intensities of economic sectors and industries, and
the absolute value of their corresponding GHG emissions. These two are associated with
the technological characteristics of their respective processes of production.
Economic models of different types deal with various aspects of Climate Change
mitigation policies, or with the same aspects but using different approaches and inbuilt
assumptions (Macroeconomic models, Econometric models, General Equilibrium models,
etc.).
The present study is in principle concerned only with those models within the Input-
Output or Structural Analysis tradition, which can be defined as meso economic models,

3

that is to say they are not macro, nor micro economic models. They deal with sectoral
economic magnitudes.
We are building and developing an Environmental Input-Output (EIO) model of the
Mexican Economy for the purpose of analysing the effects of a change in technology in
some of the key sectors of the economy identified as both strategic and highly GHG
emitters. The period in which this impact analysis is studied goes from 2008 to 2020. The
main idea is to find out to what extent the use of more efficient technologies in key
economic sectors makes the reduction of GHG emissions possible under different scenarios
of GDP growth.
It must be said from the beginning that, although most of the information we are
using here may be called hard data since it comes from official surveys and has been
subject to verification, the resulting forecasted data and the simulations only indicate
tendencies subject to assumptions and not real values, as in any other model interpreting
the economic reality.
This report is divided into seven sections including this introduction. The second
section is devoted to review the literature concerning some of the main economic aspects of
climate change and mitigation. The third section is a comparison between Mexico and
Canada. In the fourth section we determine the key or strategic sectors, and also the most
polluting sectors of the Mexican economy, using Input-Output analysis. The fifth section
deals with the Environment Input-Output model designed to estimate GHG emissions by
sector for the period 2008-2020 under various scenarios including technology changes
aimed to reduce GHG emissions. In the sixth section the outcomes of the model are
analysed. Finally, in the last section, we present some conclusions and policy
recommendations based on the results of the model.


4

II. REVIEW OF SELECTED LITERATURE

1. Climate Change, Causes and Mitigation Policies
All the IPCC reports and technical papers consulted (IPCC, 1996, 2000, 2001, 2007)
indicated, and the Stern Review (2006) confirmed, that economic activities are the sources
of GHG emissions and, therefore, represent the anthropogenic causes of Climate Change.
Even though the Climate Change problem is a global one, it is originally caused by
the GHG emissions in the production processes of specific sectors of the economy of every
country. Thus, in order to first assess the magnitude of the GHG emissions, the IPCC
defined instruments and methods for that purpose applicable to each country. Among these
methods, there is a common sector classification system. This has been useful to identify
those sectors that contribute most to generate GHG emissions in each country. There was,
however, a second consideration of equal importance and that is the GHG emissions
intensity of each sector, i.e. GHG emissions per unit of output in the respective sector. Here
we find as a challenge how to make compatible IPCC sectors classification with each
countrys National Income Accounts classification of industries and sectors in order to
estimate correctly absolute and relative GHG emissions in each economic sector or
industry. In any case, the important point to stress is that the relevant information, for
purposes of measuring and, therefore, applying policies towards abating GHG emissions is
clearly sector specific.
According to all the literature reviewed, in the sphere of production GHG some
mitigation policies are centred on the introduction of abatement technologies. In some cases
these technologies are sector-specific. In other words, in Economics the definition of an
industry (sometimes called a sector) is a group of firms that produce more or less the same
good or service and therefore share, more or less, the same technology. In practice,
however, there may be significant technological differences between firms in the same
sector, especially when the market structure is imperfectly competitive.
The abatement technologies according to the various authors reviewed can be of
two types: one is called end-of-the-pipe technology that reduces GHG emissions without
implying other changes in the production process; another is the technology that implies an
important change of the production process reflected in the input-output technical
coefficients in order to reduce GHG emissions.

5

2. Policy instruments and economic models
In order to induce the use of technology that reduces GHG emissions by the producers
(switching from a conventional technology to an abatement one), every government has a
variety of instruments and measures to apply: market based programs, regulatory measures,
voluntary agreements, scientific research and development (R&D), and infrastructural
measures. IPCC maintains the idea that there is no best single instrument or measure to
apply but rather a combination of measures adapted to national, regional and local
conditions will be required (IPCC, 1996). The same position is favoured by the OECD in
its studies.
For estimating the economic impacts of Climate Change and the future economic
scenarios, either globally or regionally, there have been built economic models of various
types, among these the most relevant are: Econometric models, General Equilibrium
models and Environmental Input-Output models. Some of these models, at country-level,
have also been used to estimate the impact on GHG emissions reduction of different
policies aimed to that purpose. These types of models have been widely recognised as valid
and useful by the IPCC and the OECD in their respective studies.
But only some of these models where favoured in the beginning for analysing
Climate Changes economic impacts. As Barker pointed out by 1998: Most of the research
effort has gone into the development of computable general equilibrium (CGE) models, and
this methodology dominates the field (Barker, 1998). This type of models is based on
Neoclassical thinking, that is, they favour market mechanism solutions over state policies
instruments, in almost any economic issue including of course, ecological ones. Therefore,
in the policy instruments utilized for ecological problems the use of market based policies
was predominant. All this has been due to the prevalence of the so called Mainstream
Economics in most countries and in the most important international financial institutions,
for over twenty five years.
Whatever the extent of market oriented policies carried out between 1988 and 2005
they did very little in solving the GHG emissions problem, called Climate Change.
Nicholas Stern pointed out in his Review, in 2006 after eighteen years of IPCC foundation,
that Climate Change was the greatest and widest-ranging market failure ever seen
(Stern, 2006). Of course he had not witnessed the so called Sub-prime financial crisis of

6

2008, initiated in the US but extended to rest of the world, which was also a market failure.
Stern (2006) also called for a major change (as opposed to a marginal one) in GHG
reductions which, as all major changes in the economy must, in our opinion, be led by the
state in each country case.
The need for state intervention arises also from the existence of market
imperfections in each and every economy in the world. It is not surprising that the OECD
emphasises that putting a price on GHG emissions through price mechanisms, has the
limitation that they do not address the full range of market imperfections that prevent
emissions to be cut at least cost, such as information problems, (Duval, 2008).
The OECD finds also that empirical analysis indicates that the most important
determinant of innovation in the area of renewable energy technologies is general
innovative capacity. According to Furman, et al. (2002), National innovative capacity is
the ability of a country to produce and commercialize a flow of innovative technology over
the long term. National innovative capacity depends on the strength of a nations common
innovation infrastructure, the environment for innovation in a nations industrial clusters,
and the strength of linkages between these two. However the OECD study says in the
case of energy public policy makes a difference. Public R&D expenditures on renewable
energies induce innovation, as do targeted measures such as renewable energy certificates
and feed-in tariffs (Hai, et al., 2010).
Finally, another issue that calls for state action is the issue of equity, namely the
extent to which the impacts of climate change or mitigation policies create or exacerbate
inequities both within and across nations and regions. This implies the need for the
application of state policy measures aiming to prevent or to compensate any inequities that
may result from either climate change impacts or mitigation policies, between sectors or
population groups within a country, and internationally agreed regulations in the same
direction for inequities between countries.
3. General Equilibrium models and Climate change.
General Equilibrium in general
General equilibrium models are rooted in Neoclassical thinking since they pay particular
attention to the specification of demand and supply functions, derived from the assumption

7

of utility and profit maximizing consumers and firms; to assume perfect competition; and to
impose market clearing (Mercado, 2003).
It is said, from the conventional neoclassical approach, that Economics is about
choice subject to constraints. Under this approach, the main problem for Economics is to
decide upon the allocation of resources in order to maximize Social Welfare which is a
function of the addition of the welfare functions of all individuals in society, that is, their
utility functions. This optimization problem is subject to three sets of constraints related to:
factor endowments, technology and tastes. Optimization takes place in consumption,
production and product-mix. This process looks for the most efficient solution in Paretos
terminology. The solution is a set of prices for goods and factor services satisfying
simultaneously all the equations in the various markets. The only possible way that a free
market economy finds the most efficient solution rests in four basic assumptions: perfect
competitive markets, constant returns to scale for every firm in the market, that there are no
externalities and that there is no market failure connected to uncertainty (Layard and
Walters, 1978, Chapter I). Another important assumption not always explicit is full
employment of all factors of production. This, in fact, is what General Equilibrium is all
about.
Most of the assumptions, implicit or explicit, in a General Equilibrium model are so
unrealistic that make it difficult to apply to the real world economy. In general there is a
high probability that many real world economic issues cannot be solved by free market
forces working by themselves but require state action, that is, a state or public policy. This
situation is called market failure. Be it due to market imperfections, the presence of
externalities, the existence of monopoly rents from increasing returns to scale, or the
existence of uncertainties.
Computable General Equilibrium models
Computable General Equilibrium (CGE) models are those models almost exactly as the
above mentioned, in this case trying to use information taken from real economic data, but
keeping the usual assumptions, which become actually very strong limitations. They try to
capture a wide range of economy interactions between a variety of economic agents and
institutions. Given some behavioural assumptions with respect to those agents and
institutions and with respect to the functioning of markets, these models are used to

8

determine relative prices and quantities produced and consumed. They sometimes provide a
relatively high disaggregated picture of the economy.
A CGE model requires computational techniques to be solved. Many times the
parameters and variable values of these models are estimated with information obtained
from a Social Accounting Matrix (SAM) which contains information on the flow of goods
and payments between institutions in the economy. This is the only real link between these
models and the Input-Output models, since SAM and Input-Output data come from the
same source: the National Income Accounts of a country.
CGE models have traditionally been used to answer comparative static or what if
questions, that is, they are mostly static. However, dynamic specifications are been
increasingly used for forecasting purposes (Dixon and Rimmer, 2009).
CGE models for Climate Change
There is a wide variety of CGE models dealing with Climate Change problematic, some of
them using Input-Output Tables, Social Account Matrices, and/or Econometric equations to
forecasting specific variables and some of them are considered hybrid models because they
deal, simultaneously, with economic and physical data of Climate Change. Therefore, there
is sometimes a combination of different techniques in the same model.
There are CGE studies for a region or area within a country (see for instance Rose,
et al., 2000); there are others for only one sector in a given economy (see for instance Zhai,
et al., 2009); others are for a single country (see Fullerton and Heutel, 2010; Dejuan, et al.,
2008), for various world regions, or for the whole world (see for instance, Ross, 2008;
Sassi, et al., 2010). The topics treated vary, there are: sector analyses, energy demand
estimations, Climate Change policy analyses, sustainable economic growth policies. In all
of these cases the idea is to reduce the expected economic costs or impacts of Climate
Change, choosing the right (more efficient) Climate Change mitigation and/or adaptation
policies. These costs are generally measured in terms of welfare gains or losses for present
and/or of future consumers.
Nordhaus GE models
Although there seems to be several important General Equilibrium models dealing with
various aspects of Climate Change, the most cited in the literature are those of William D.
Nordhaus from the University of Yale in New Haven, USA.

9

Nordhaus modelling on energy can be traced as far back as 1973 (Nordhaus, 1973).
But it seems that the initial relevant work done in the seventies was in his energy model of
1979 for the US energy sector where he tries to determine the prices of energy resources,
for an efficient use of those resources (called efficient prices). The investigation is
oriented towards establishing the time pattern of the efficient use of the energy resources
assuming that those resources which are scarce have a royalty attached to it that increases
over time with the market interest rate. The difficulties the study finds in trying to adapt
economic theory to real world facts, for instance the assumption of competitive oil markets
that yield competitive oil prices versus actual oil prices determined by some degree of
monopoly in the real oil market, leads the investigation to formulate the actual question of
what is the chance that global environmental effects will appear as a result of unrestricted
market forces?. In answering this question Nordhaus concludes that we are probably
heading for major climatic changes over the next 200 years if market forces are
unchecked, He therefore propose a carbon tax as the most efficient control strategy
(Nordhaus, 1979). The existence of non-competitive markets brings about some degree of
uncertainty which adds to that inherent to the costs of new technologies estimates. It is
therefore recognized that the validity of the results in this type of models is restricted by the
very optimistic assumptions that there are no significant impediments for the action of
market forces (Nordhaus, 1979).
In 1983, Nordhaus and Yohe presented a world probabilistic model for estimating
CO
2
emissions as influenced by major uncertain variables or parameters. The technique
utilized is called probabilistic scenario analysis. The model is a highly aggregated model
of the world economy and energy sector. The main equation is a multi-input production that
related Gross National Product to labour, fossil fuels and non-fossil fuels inputs. The so
called key uncertainties included in the model are, the rate of population growth, the
availability and cost of fossil fuels, the productivity growth rate, and some others. The
important findings in this model are odds are even whether the doubling of carbon dioxide
will occur in the period 2050-2100 or outside that period it is a 1-in-4 possibility that
CO
2
doubling will occur before 2050 and 1-in-20 possibility that doubling will occur before
2035 (Nordhaus and Yohe, 1983, p. 94).

10

Nordhaus DICE model is presented in 1992 (Nordhaus, 1992). It is called DICE
for a Dynamic Integrated Climate Economy model. This model is said to correct for the
shortcomings of previous studies. The basic idea is to use a Ramsey model of optimal
economic growth with certain adjustments and to calculate the optimal path for both capital
accumulation and GHG emissions reductions.
The model is an optimal-growth model for the world economy. It is designed to
maximize the discounted utility or satisfaction from consumption subject to a number of
economic and climatic constraints. The global economy is assumed to produce a composite
commodity. The composite economy is endowed with initial stock of capital and labour and
an initial level of technology and all industries behave competitively. Each country
maximizes an inter-temporal objective function identical in each region which is the sum of
discounted utilities. Population growth and technological change are exogenous. There is
no need for international trade since the outputs of the different countries are perfect
substitutes.
Another important feature of this model is that it is assumed that GHG emissions
can be controlled by increasing the prices of factors or outputs that are GHG-intensive.
The presentation also says that the model can be interpreted either as an optimizing
framework or as an outcome of idealized competitive markets. It is assumed that the public
goods nature of climate change is somehow overcome in an efficient manner. That is, it
assumes that, through some mechanism, countries internalize, in their decision making, the
global costs of their emissions decisions.
Two subjective variables which are not discussed, seem to be important in the
model: in the Objective function, defined as the pure rate of social time preference later
introduced as the market rate of interest or the marginal productivity of capital, and g
in the Utility equation defined as inequality aversion which is assumed equal to zero for
no apparent reason.
We emphasize on the importance of these two parameters, for the following
reasons: plays an important role in determining the discounted value of utility, which
later would become a major issue in Nordhaus criticisms of the Stern Review (Nordhaus,
2006); in turn g is related to the Atkinson Index (Atkinson,1970) which captures subjective
inequality. The Atkinson index depends on the inequality aversion parameter, e, of the

11

decision-maker or society, and measures the fraction of total income which could be given
up with no loss of social welfare if the remainder were to be distributed equally (Lambert
et al., 2003). While Nordhaus assumed this parameter as equal to zero in his growth model,
Stern (1977) found e = 1.97 for the UK income tax code in fiscal year 1973/4 Sterns
paper also contains a comprehensive survey of many other approaches to evaluating the
elasticities of both private and social marginal utilities of income, reporting values found by
a range of authors for different countries using various methodologies as high as e = 10 and
as low as e = 0.4 (Lambert et al., 2003).
The important conclusions from this version of Nordhaus model results are that an
efficient strategy for coping with greenhouse warming must weigh the costs and benefits of
different policies at different points of timeEstimates of both costs and damages are
highly uncertain and incompleteIn terms of damages the impact of climate change
coming from a 3C rise in global mean surface temperatureis estimated to be a about 1.3
of output for the global economy (Nordhaus, 1992).
As an improvement of the DICE model, a new model called RICE is presented in
1996, by Nordhaus and Yang. The name stands for Regional Integrated model of Climate
and the Economy. This is described as a regional dynamic general equilibrium model of the
economy which integrates economic activity with the sources emissions and consequences
of greenhouse-gas emissions and climate change. By disaggregating into countries the
model analyses different national strategies in climate change policy. The model asks how
nations would in practice choose climate-change policies in light of economic trade-offs
and national self-interests for reductions of GHGs. In the RICE model the world is divided
into 10 regions, each is endowed with an initial capital stock, population, and technology.
Of these three variables capital accumulation is determined by optimizing the flow of
consumption over time. The major economic choices faced by nations (or the group of
nations) are: (a) to consume goods and services; (b) to invest in productive capital and (c)
to slow climate change through reducing CO
2
emissions. In the model there are also three
types of strategies undertaken by nations to deal with GHG emissions: (1) Market policies,
meaning no controls on GHG emissions; (2) Cooperative policies, nations agree to reduce
CO2 emissions in a globally efficient way and (3) Non cooperative policies, individual
nations undertake policies that are in their national self-interests ignoring the spill overs of

12

their actions to other nations. From to the results of the model there are seven basic
conclusions. The most important seems that the model estimates the difference between
cooperative efficient policy and the non-cooperative policy. This latter is one in which
countries maximize their economic welfare taking policies of other countries as given.
This implies that small countries whose climate change policies have little effect on their
own economic welfare, will have little incentive to reduce emissions while the largest
countries will have greatly attenuated incentives to engage in costly reductions in CO
2

emissions (Nordhaus, 1996). The results of the model indicate that the stakes in
controlling global warming are modest in the context of overall economic activity over the
next century. The estimates indicate that losses from global warming will be in the range of
1 to 2 per cent of global income over the next century. According to the model successful
cooperation would lead to net gains, but the failure to cooperate is unlikely to lead to
economic disaster over the next century.
In a book published in a digital and paper version as well, called Roll the DICE
again: Economic Models of Global Warming by Nordhaus and Boyer (1999) the authors
made a detailed description of Nordhaus general equilibrium world models built until then
and they run a new version of DICE. In Chapter 4 of this book the impacts of climate
change are analysed. The model called RICE-99 estimates damage functions for both the
world and by region and sector. The results seem to be of the greatest importance. The
chapter says in page 31: The results differ markedly by region. The impacts (of a 2.5C
global warming) range, from a net benefit of 0.7 per cent of output, for Russia, to a net
damage of almost 5 per cent, for India. The global average impact is estimated to be 1.5 per
cent of output, using projected output weights and 1.9 per cent of output using 1995
regional population weights. Current projections of RICE-99 indicate that total warming
in an uncontrolled environment will be slightly below 2.5C around 2100. Our estimate is
that damages are likely to be around 1.9 per cent of global income using 2100 output
weights. The damages for the US, Japan Russia and China are essentially zero over that
time frame, assuming that catastrophic scenarios do not materialize. Europe, India and
many low income regions appear vulnerable to significant damages over the next century.
Right after his debate with Stern, Nordhaus published a new book The Challenge of
Global Warming: Economic Models and Environmental Policy (Nordhaus, 2007) with new

13

or revised projections and estimations of his RICE and DICE models. The models are again
General Equilibrium models, with the usual assumptions for general equilibrium and the
emphasis on welfare economics considerations as central for the evaluation of policies. In
this case the central questions to be answered are: How sharply should countries reduce
CO
2
and other GHG emissions? Should there be a system of emissions limits imposed on
firms, industries and nations? Or should emissions reductions be primarily imposed through
taxes on GHG?
The author of the study begins right away to answer these questions: In practice an
economic analysis of climate change weights the costs of slowing climate change against
the damages of more rapid climate change. Economic history and analysis indicate that it
will be most effective to use market signals, primarily higher prices of carbon fuels, to give
signals and provide incentives for consumers and firms to change their energy use and
reduce their carbon emissions. In the longer run, higher carbon prices will provide
incentives for firms to ease the transition to a low-carbon future.
The major results of the model are: The base line case projects a rapid and
continued increase in CO2 emissions by 2100, which will increase the mean global surface
temperature by 3.1C by 2100 and 5.3C by 2200, relative to 1900. Climate changes are
estimated to increase global damages by 3 per cent of global output in 2100 and close to 8
per cent of global output in 2200.
As in previous studies an important result is the estimation of the optimal carbon
price or optimal carbon tax, since it is the policy most favoured by this type of analysis. So
it is called the efficient policy.
General Equilibrium Environment models in Mexico
Although in Mexico there have been many General Equilibrium models dealing with
various aspects of the economy. The only known GE models that we know deal with
environmental issues are the one by Roy Boyd and Maria E. Ibarrarn (Boyd and Ibarrarn,
2008) dealing with one of the probable effects of climate change drought in Mexico; the
other is called Anlisis econmico robusto para el desarrollo de estrategias de bajas
emisiones para Mxico (A robust economic analysis for the development of low emissions
strategies for Mexico), just recently produced by Hctor M. Bravo Prez, Juan C. Castro

14

Ramrez y Miguel A. Gutirrez Andrade, all staff members of the Faculty of Economics at
the University of Mexico (UNAM), still unpublished.
The authors of the first study mentioned (Boyd and Ibarrarn) state in the articles
abstract the following: Climate change is increasing the intensity of extreme weather
events. Mexico is particularly prone to suffer at least two different types of these events:
droughts and hurricanes. This paper focuses on the effects of an extended drought on the
Mexican economy. Through a computable general equilibrium model, we simulate the
impact of a drought that affects primarily agriculture, livestock, forestry, and hydropower
generation. We look at the effects on the overall economy. We then simulate the effects of
several adaptation strategies in (chiefly) the agricultural, forestry, and power sectors, and
we arrive at some tentative yet significant conclusions. We find that the effects of such an
event vary substantially by sector with moderate to severe overall impacts. Furthermore, we
find that adaptation policies can only effect modest changes to the economic losses to be
suffered.
The second study longer and maybe more ambitious has a summary that goes like
this (translated from Spanish):
The work has the objective of calculating the effect on the distribution of income in
Mexico of a tax applied on fossil fuels demand with the purpose of reducing CO
2

emissions. To this end two Computable General Equilibrium models are built, one is static
the other dynamic. The methodology followed is the one proposed by Shoven and Whalley
and the software used is GAMS. There are various taxes according to the different fossils
fuels types and there is also one general tax to all. There is a simulation of the economic
behaviour in two scenarios according to the value of the elasticity of energy substitution: a
rigid elasticity of 0.2 and a flexible of 20. The Mexican economy turned out to be closest to
rigid elasticity values. The results of the model show that the tax strategy has the desired
effect of reducing the fossil fuel demand, but taxing coal affect only its demand not
generating a substitution effect towards any other energy good. On the other hand the
welfare effect is clearly differentiated. It depends on the relative importance coal play in the
household as an input in home meals preparation. As a consequence of this simulated tax,
households are divided into two groups. The distributive effects are negligible given the
small importance coal has, compared to other energy goods in the economy. But for other

15

energy goods different from coal, there are substitution and complementary effects when a
tax is applied to these goods. In this case there is a high income concentration effect. For a
flexible economy, welfare losses are lower than for a rigid economy. Therefore, the most
important conclusion is that in a rigid economy substitution between energy goods is more
difficult and produces a welfare loss. Thus, a public policy aimed to reduce the social costs
of reducing CO
2
emissions by way of taxes, must take into account the type of fossil energy
good to be taxed and besides it should try to induce technological change and/or change the
type of current regulation rules for energy substitution, that is subsidies or taxes to energy
products (Bravo, H. et al., 2012).
4. Econometric models
With respect to applied models we reviewed only two types: Macro-econometric models
and Environment Input-Output models. Among the first group there is one world-wide
model called the Energy-Environment-Economy Model at the Global level (E3MG)
elaborated by experts from the University of Cambridge, and which in combination with a
chemistry transport model called p-TOMCAT has also been applied to the case of Mexico.
The E3MG model is highly appealing to us, for two main reasons: (1) It is based on
historical data collected by official agencies; (2) It is defined as Post Keynesian, in the
sense that, it is demand driven and it does not assume full employment or perfect
competition in the neoclassical fashion. Also because it takes into account various
economic sectors, that is, it is defined as a multi-sector model (Barker, et al., 2008).
The model applied to Mexico by Terry Barker and his group provided an important
insight on the situation and perspectives of Climate Change impacts and mitigation options
for Mexico. The results show that if Mexicos government applies policy measures oriented
to reduce fossil fuel consumption scenario called low carbon Mexico it can improve its
rate of growth in the medium and long run from 3.61 per cent a year 2005-2050 in real
terms (baseline trend) to 3.64 per cent, (0.03 percentage points increase), reducing at the
same time GHG emissions by 80 per cent in 2050 with respect to the trend. The strategy is
clear and straightforward: Stronger regulations help to upgrade the vehicle stock towards
low or zero emission vehicles, switch the power sector substantially to renewables and
improve energy efficiency in industry and buildings. These are complemented by a carbon-
emissions trading scheme, to reduce emissions from the energy sectors, and carbon taxes on

16

other sectors (Barker, et. al., 2010). The results could be improved if the rest of the world
follows a similar low-carbon strategy at the same time. In such a case, Mexicos annual
growth rate would be increased even further (to 3.65 per cent a year 2005-2050).
The other model applied to Mexico (SEMARNAT, 2009a) is also a macro-
econometric and multi-sectoral model. Its results, however, are not as optimistic as Barker
models. The argument of SEMARNAT goes like this: if we assume a mean growth rate of
3.5 per cent of GDP in real terms a year for the Mexican economy and do nothing for
mitigation, then the desired 50 per cent reduction of GHG emissions by 2050 (with respect
to 2002) cannot be accomplished. In order to reach this reduction target we have to incur in
various mitigation costs that amount to an average of 2 percentage points in the annual
level of GDP. So the economy cannot grow as fast, due to high mitigation costs.
In that respect Barker argues that: such results are largely based on equilibrium
modelling in which the cost increases are assumed in the theory underlying the models, and
technological change is assumed to be exogenous. He also explains that: the difference in
results for GDP comes from the assumptions that revenues [in his model] are recycled to
fund the low-carbon investment, and that there are underemployed resources in the
economy to allow for faster growth. If we assume that the extra investment crowds out
private consumption then the model gives a reduction in GDP of 0.5 per cent by 2050.
5. Applied Environmental Input-Output models
Since Leontiefs important works on environmental issues (Leontief, 1970, 1973), Input-
Output models have incorporated into the analysis, pollution and pollution elimination
activities, for instance in modern Input-Output textbooks (Miller and Blair, 2009; Ten Raa,
2005) and in a wide variety of published books, articles and chapters in books, some of
them cited in the Bibliography.
In some of the various studies and reports dealing with Climate Change and
technology, published by Intergovernmental Panel on Climate Change (IPCC) (1996, 2001)
and the Organization for Economic Cooperation and Development (OECD) (Hai et al.,
2010, Duval, 2008), Input-Output models are considered useful and valid methods for
estimating economic costs and impacts of climate change and its abatement technologies.
The same consideration is given by Terry Barker (see Barker, 1998).

17

With respect to I-O models specifically relating to climate change, mitigation
technologies, impacts and costs evaluation at a national level, we have found a very limited
number of references. Of them all, there are four that called our attention for various
reasons: they were applied to a country not a region within a country, or a region consisting
of several countries; they explicitely dealt with the application of abatement technologies
for reducing GHG emissions in the Input-Output matrix of that country; the models were
explicitely formulated; they were published and/or easy accesible; and finally there were
actually applied and produced results in a practical case. These were the models applied to
the Dutch economy: Idenburg (1998); Idenburg and Wilting (2000, 2004); Wilting, Faber
and Idenburg (2004); Brink and Idenburg (2007).
The EIO models applied to the Netherlands are in fact two different ones, with
different characteristics and objectives. The first one in time was DIMITRI. This was
theoretically developed by Idenburg (1998) mostly based on Duchin and Langes model
regarding technological change and Input-Output (1992, 1994) and explicitly aimed to
evaluate the impacts of the implementation by the Dutch government of GHG emissions
abatement technologies on the Environment, and Health and Welfare of Dutch population.
However, the model so designed could not be used with empirical data until few years later
in a second paper (Idenburg and Wilting, 2000, 2004). In this paper the authors claim that
they developed a dynamic input-output model of the Dutch economy that enabled the
investigation of effects of technical changes in individual economic sectors on the whole
economy and the environment.
A crucial aspect for the dynamics and the introduction of new technologies in the
model are the variables and equations regarding: investment by sector, capital goods
capacity (existing, expected and planned) by sector, depreciation rates by sector, and the so
called Matrix of capital coefficients. The installed technology is a mix of technologies
implemented in previous periods. As a result of depreciation and new investments, the
installed technology in all sectors changes every period. After installing new technologies,
the technological matrix depicts the new installed mix of technologies. The model estimates
the technological matrices for each period.
They applied the model so constructed to the Dutch economy for the period 1980-97
trying to interpret the actual technological change occurred with the model and then tested

18

its estimated results to actual values, finding no statistically significant discrepancy
between them. The authors compared also the models outcomes to a set of imaginary
projected data assuming that no technological change had occurred, that is, assuming fixed
1980 technology. The results of this comparison are somehow contradictory. While in the
sector of business services production changing technology caused a higher growth rate
than fixed technology, in agriculture it was the opposite. With respect to energy use the
results show the opposite situation in each of these two sectors, namely, in business
services there is an increase in energy use by the changing technology and in agriculture
there is a decrease, both compared to fixed imaginary technology. But the authors claim
that the exercise was just to validate the use of DIMITRI for investigating the effects on
new technologies in future scenarios. And to do that they say a technology database
including at least a new available technology for each sector is required.
The third application of DIMITRI to the Dutch economy reviewed is the one
presented in the paper by Wilting, Faber and Idenburg (2004). In this case the idea was to
explore possible scenarios of technological change aimed to reducing GHG emissions.
This paper presents a method that combines an extrapolation method of technical
coefficients with more specific knowledge on technologies. The extrapolation method
generates a reference path, independent of the scenarios. This reference path can be seen as
a more or less autonomous path to the future, based on the past trends. The scenarios are
variations based on this reference path. For the specific technology scenarios, detailed
information on the rise and fall of technologies is implemented. In the reference scenario,
technical coefficients are based on historical trends.
The method was applied in four scenarios, which are based on the framework
developed by the IPCC (2000). A large number of climate change scenarios are clustered
into four which are referred to as A1, B1, A2 and B2. These scenarios are descriptive in
nature, distinguished from each other along two lines: from efficiency to equity; from
globalization to regionalization.
The production calculated for the four IPCC scenarios fluctuates around the values
in the reference scenario. For almost all sectors the production in the A1-scenario is the
highest. This scenario shows a further increase in production in business services, which

19

leads to an increase of production in other sectors. In all scenarios electricity production
increases, due to a higher rate of electrification in the future.
The emissions in the A-scenarios are lower due to a relative high share of genetic
modification techniques. Organic agricultural techniques will be partly applied in the B2-
scenario (and to a lesser extent in the B1 scenario) in 2030. In A1, A2 and B1 scenarios, a
further decrease in nitrogen emissions is achieved as a result from some general efficiency
improvements. Finally, the introduction of alternative agricultural production chains results
in fewer emissions especially in B1 and to a lesser extent in A1 and B2.
The second type of model applied to the Dutch economy is the one developed by
Brink and Idenburg (2007). It is different from DIMITRI in two important features: (1) it is
not dynamic, so it does not include any explicit data or estimation of capital stock,
investment, depreciation by sector, nor a capital coefficients matrix; (2) it is built with the
purpose of studying the effects of the application of the best GHG abatement technology
per sector, choosing one among various alternatives. The selection is based on a total cost
analysis implemented in an optimization I-O model. The technologies considered for
election are all add-on technologies, that is, they do not imply a change in the product or in
the production process.
An important feature of this model is that it assumes the working of a permit
scheme under free market rules. That is to say the model privileges an environmental policy
of GHG reduction through a permit market system.
The model is said to be based on a modified Leontief EIO model, which is, in turn,
extended for cost effective analysis. The model is extended by including q abatement
technologies that can be applied to reduce the emissions of k pollutants. Abatement
technologies are included as separate production processes that can be added to the
production processes producing the commodities. The total reduction in emissions is the
sum of the reduction by the separate technologies. The cost of abatement depends on the
technologies available and their cost. The cost of abatement is made up of the amount of
inputs from other sectors and the primary inputs (labour) required for using the abatement
technologies. The price of emission permits depends on the quantity of permits available
and the abatement cost of the various trading partners.

20

The model is formulated as an optimization problem minimizing total cost of
production (i.e. the sum of the cost of the primary inputs in the various sectors) given
restrictions on final demand and quantities of emission permits. Decision variables are the
production level in the various sectors, the level of abatement by the various abatement
technologies and the amount of permits obtained by each sector.
Abatement technologies are aimed to the reduction of one pollutant. The price of
emission permits for a specific pollutant paid by a sector cannot be lower than the marginal
cost of reduction of that pollutant in the same sector. The price of commodities is affected
by environmental policy through the price of emissions which depends on the permit price
and the cost of abatement. This implies that with more stringent restrictions on the quantity
of emission permits, resulting in a higher permit prices and marginal abatement costs, the
cost of polluting production increases and hence the price of the associated commodities.
The authors used a highly aggregated Input-Output table of the Dutch economy,
with five production sectors causing emissions of two pollutants. Emissions can be abated
by a number of abatement technologies with different abatement potential and different
costs. If a sector faces a price for emission permits that is higher than the price of abatement
by a certain abatement option in that sector, the sector will implement this abatement
option. The price of the permits will at least be as high as the price of abatement by the
most expensive abatement option that is implemented in the economy.
Since there is free trade of permits among sectors, permits will be allocated over
sectors in such a way that the total abatement cost over all sectors is minimized. As a
consequence of the system of tradable emission permits, emissions will have a price that
adds to the total production cost. Results of calculations with the model show that a cost-
efficient overall reduction of CO
2
emissions by up to 5 per cent can be realized at relatively
low cost by abatement in the agriculture and industry. A further reduction in emissions (up
to 20 per cent) requires substantial abatement costs, in particular for abatement options in
the energy sector. Total abatement cost in the energy sector increase to more than 7 per cent
of value added.
These two types of EIO models applied to the Dutch economy have both different
purposes and therefore different assumptions: DIMITRI is dynamic and Brink and
Idenburgs (B&I) is static. DIMITRI considers technological change as a change in each

21

sectors production function, while B&Is model considers end-of-the-pipe technologies.
The first is looking for effects of technical change in various scenarios (it is a forecast
model). The second is looking for the optimal technology in terms of sectoral costs (it is an
optimization model). Both deal with Input-Output matrices highly aggregated.
The different characteristics of these models and the last similarity just mentioned
may represent a disadvantage for a richer analysis in each of them, as rightfully expressed
by B&I (2007). However, we consider both models results very important for policy-
making and technically worthy. As always the major difficulty is the availability,
opportunity and reliability of the information needed to apply them to another countrys
case.


22

III. A COMPARISON MEXICO-CANADA

In a previous study (SEMARNAT, 2009b) we analysed the 2003, Mexican Input-Output
matrix and identified the most pollutant sectors, both in absolute and relative terms,
according to GHG emissions inventory reported by Mexican official agencies. We also
made a comparison of our estimates to another countrys for which we chose Canada, a
nation which has been close to Mexico especially since the operation of the North
American Free Trade Agreement (NAFTA) in 1994. Canada is among those countries
engaged in implementing GHG emissions reduction policies, as required by the Kyoto
Protocol.
Canada has had a complete follow up of GHG emissions reduction by sector and of
the costs incurred by firms for GHG emissions reduction associated in particular to their
respective production levels. With this information and Canadas GDP by sector we have
estimated current and capital expenses realized per sector for GHG emissions reduction.
We use this information to estimate the corresponding costs for the same sectors of the
Mexican economy, as if these Canadian abatement technologies were applied in Mexico.
1. Method for estimating comparison indicators
We gathered annual data about current and capital expenses incurred to reducing GHG by
sector in Canada this was available for the years 2002 and 2004. We also obtained the
actual levels of GHG emissions for the years 2003 and 2005. This allowed us to assume a
period of adaptation for new investments to actually reduce GHG emissions. With this
information we calculated the cost of reduction of GHG per ton of CO
2
equivalent. In turn
this allowed the comparison with the same sectors in the Mexican economy and the
estimation of their corresponding GHG emissions reduction in these selected sectors.
There were fifteen industries for which this information was available in Canada all
of which belong to the energy sector in its stationary and industrial subsectors, according to
the IPCC source classification. They emitted in the whole 45 and 38 per cent of total of
GHG emissions for the years 2003 and 2005. These 15 industries represented more than 20
per cent of Canadas GDP, in the economy 79 industries. The only important industries
missing were those related to transportation, that in many countries are high GHG
emissions producers. But, at that time, there was no information about any GHG reduction
activities and their corresponding costs for these industries. It is important to stress the fact

23

that Canada reduced in an important way the level of GHG emissions intensity relative to
GDP in the period analysed.
In Mexico all the industries considered for comparison were for the same period
high GHG emissions producers; they represented 57 per cent of total GHG emissions and
their share of the GDP was also of more than 20 per cent. The only industry which is not
even close as important as that of Canada is Pipeline Transportation.
2. Estimation results
The results of the calculations and comparison are shown in Tables 1 and 2. The first is
related to Canada and indicates that with a total expense equivalent to 0.5 per cent of GDP
of these industries, GHG emissions were reduced in 6 per cent per year. It must be said,
however, that we did not apply an I-O analysis or a macro dynamic analysis. So, there are
no feedback and multiplier effects of these investments on the Canadian economy across
time being considered here.
The results varied among industries, in some cases GHG emissions variations were
not negative as expected but positive. This might be attributed to the different periods the
various investments take to start producing positive results in different industries, also to
the fact that different industries have different rates of growth.
There is an important case that must be mentioned, that of industry (3) Forestry, that
switched from high CO
2
emitter to important CO
2
capturer. The outstanding performance
of this industry is due to the very successful reforestation policy followed by Canada, at
very low cost per GHG ton relative to the rest of the industries.
For Mexico the estimation of GHG emission intensities with respect to GDP, as
compared to Canadas indicates that in that period and in the same group of industries,
Mexico was in general more inefficient than Canada, that is to say, each unit of product
generated as a by-product a higher GHG emission quantity in Mexico than in Canada.
But probably the most important result is that by applying the same GHG abatement
technical methods used in Canada in each of the selected industries and assuming the same
Canadian costs converted into Mexican pesos Mexico would have to expend 8.5 per
cent of these industries GDP in order to eliminate, or reduce to zero, the GHG emissions of
these industries, which represented 57 per cent of the total GHG emissions.

24

Someone may say that these comparisons Mexico-Canada may not be totally valid
due to the important differences existing between these two countries regarding the specific
characteristics of each countrys climate, economic structure and institutional setup. So, the
Canadian abatement methods could not simply be transposed into the Mexican context.
This is true to some extent. However, Mexico and Canada have the same technical
coefficient matrix structure with, of course, different numbers in each. If we understand a
sectors technology as a function of production represented by the sectors column of the
technical coefficient matrix and we are referring to the same sectors in Mexico and Canada
we can compare technologies and we can even substitute them one for the other. Assuming
of course, everything else remains the same.


25

Table 1
Expenses on GHG Cost per
No. Industry emissions reductionTon of GHG Var.
2003 2005 Variation Mill Dlls % GDP reduced 2003 2005
3 Forestry and logging 11,000 -27,000 -19,000 45.5 0.76 2.39 2.03 -4.43 -3.23
6 Oil and gas extraction 69,164 68,244 -460 181.6 0.38 394.64 1.22 0.78 -0.22
7 Metal ores and mineral mining 15,700 15,600 -50 38.0 0.35 760.00 1.69 0.98 -0.36
9 Electric power generation, transmission, and dist. 135,000 129,000 -3,000 172.1 0.71 57.35 5.20 4.66 -0.27
14 Food manufacturing 8,729 9,517 394 34.3 0.19 34.30 0.48 0.49 0.00
15 Beverage and tobacco manufacturing 2,988 3,399 205 6.8 0.11 6.80 0.48 0.49 0.00
20 Wood product manufacturing 5,538 6,048 255 146.5 1.12 146.50 0.48 0.49 0.00
21 Pulp, paper, and paperboard mills and paper products 8,990 7,340 -825 200.2 1.74 242.67 0.84 0.68 -0.08
23 Petroleum and coal products manufacturing 4,836 4,756 -40 33.3 0.82 833.33 1.22 1.02 -0.10
24 Chemical product and preparation manufacturing 13,210 14,250 520 91.6 0.61 91.60 0.87 0.97 0.05
26 Nonmetallic mineral product manufacturing 13,180 14,080 450 23.7 0.44 23.65 2.41 2.35 -0.03
27 Iron and steel mills and manufacturing 6,370 6,520 75 50.0 0.43 50.00 0.60 0.51 -0.05
28 Fabricated metal product manufacturing 17,200 16,200 -500 32.5 0.23 65.00 1.23 1.10 -0.07
32 Transportation equipment manufacturing 14,546 12,936 -805 29.6 0.10 36.78 0.48 0.49 0.00
41 Pipeline transportation 9,110 10,100 495 23.8 0.44 23.80 1.64 1.82 0.09
Selected Industries Total 335,560 290,990 -22,285 1,109.3 0.49 49.78 1.46 1.07 -0.20
CANADA GHG EMISSIONS AND GHG REDUCTION EXPENSES
GHG Emissions GHG Emission Intensity
Gg CO2 eq. GgCO2/GDP

26

Table 2
GHG Emission Cost per
No. Selected Industries Emissions intensity GHG Ton % of
Pesos Can Dls. GgCO2eq GHG/GVA reduced Can Dls. Pesos GVA
3 Forestry and logging 15,513.2 2,005.5 51,500 25.68 2.39 123.2 952.9 6.14
6 Oil and gas extraction 369,934.5 47,824.2 37,253 0.78 394.64 14,701.7 113,721.8 30.74
7 Metal ores and nonmetallic mineral mining 35,923.4 4,644.1 2,637 0.57 760.00 2,004.2 15,503.4 43.16
9 Electric power generation, transmission, and dist. 77,012.1 9,955.9 120,845 12.14 57.35 6,930.4 53,609.1 69.61
14 Food manufacturing 301,297.5 38,951.0 7,052 0.18 34.30 241.9 1,871.0 0.62
15 Beverage and tobacco manufacturing 69,672.6 9,007.1 1,374 0.15 6.80 9.3 72.3 0.10
20 Wood product manufacturing 17,392.5 2,248.5 551 0.25 146.50 80.7 624.4 3.59
21 Pulp, paper, and paperboard mills and paper products 27,740.5 3,586.2 1,644 0.46 242.67 398.9 3,085.4 11.12
23 Petroleum and coal products manufacturing 33,838.5 4,374.6 36,941 8.44 833.33 30,783.8 238,121.6 703.70
24 Chemical product and preparation manufacturing 116,740.0 15,091.9 8,476 0.56 91.60 776.4 6,005.6 5.14
26 Nonmetallic mineral product manufacturing 84,604.7 10,937.5 40,157 3.67 23.65 949.7 7,346.3 8.68
27 Iron and steel mills and manufacturing 76,476.1 9,886.6 11,080 1.12 50.00 554.0 4,285.2 5.60
28 Fabricated metal product manufacturing 40,461.0 5,230.7 1,282 0.25 65.00 83.3 644.5 1.59
32 Transportation equipment manufacturing 191,874.8 24,805.1 209 0.01 36.78 7.7 59.5 0.03
41 Pipeline transportation 4,999.6 646.3 0 0.00 23.80 0.0 0.0 0.00
Selected Industries Total 1,463,481.1 189,195.1 321,000 1.70 49.78 15,978.7 123,600.0 8.45
MEXICO GHG EMISSIONS AND ESTIMATED REDUCTION COSTS
Gross Value Added Total Costs of Reduction
Millions Millons

27

IV. STRATEGIC AND HIGH GHG SECTORS IN THE MEXICAN ECONOMY

In the first part of this section we use various Input-Output techniques to identify those
industries of the Mexican economy that may be called strategic from a structural point of
view. In the second, we measure GHG emissions by industry and identify those considered
as highly emitters both in relative and absolute terms. In this section of the study we are
using the National Symmetric Input-Output Matrix (IOM) of Mexico for the year 2003 and
an estimate for 2008. This matrix is presented in two formats: the domestic requirements
matrix and the total requirement matrix (including imported requirements). For this analysis
we utilize the total requirement matrix. For GHG emissions we are using the Inventario
Nacional de Gases de Efecto Invernadero (National Inventory Report of GHG), produced
by the Mexican Instituto Nacional de Ecologa (INE). In Table 1, bellow there is a
description of each one of the 79 industries of the IOM of Mexico and its classification
number in the North American Industry Classification System (NAICS). This is a similar
analysis, of the previous one we did for SEMARNAT, but this time with the most recent
information, and with different results.
1. Strategic or Key economic sectors
In Input-Output analysis, sectors or industries are labelled as strategic or key, due to their
effects on others, either through demand or through supply. The relation between any two
industries is called linkage, so we have forward linkages, those related to supply, and
backward linkages, those related to demand. We first need to find out the existence of
linkages between industries and, in each case, its relative importance. So, those industries
that have many linkages with others and these linkages are very strong, will transmit
backwardly or forwardly economic effects to others. These industries are then called
strategic or key. The reason they are called this way is that the increase or decrease in their
production, may cause a demand pull and/or a supply push variations to other industries
with effects on overall gross production, input consumption, and/or labour employment. So
these industries are essential for any growth promoting policy.
We make use of two basic indicators that allow us to evaluate the relative
importance of all industries and classify them according to their capacity to transmit
economic impulses through the system of quantities and prices that represents the IOM.

28

This capacity of disseminating impulses among industries, show their potential to generate
external diseconomies, as for instance, ecological ones.

Table 3
No. NAICS Industry
1 111 Crop production
2 112 Animal production
3 113 Forestry and logging
4 114 Fishing, hunting and trapping
5 115 Support activities for agriculture and forestry
6 211 Oil and gas extraction
7 212 Metal ores and nonmetallic mineral mining
8 213 Support activities for mining
9 221 Electric power generation, transmission, and distribution
10 222 Water, sewage and natural gas distribution
11 236 New nonresidential construction
12 237 New residential construction
13 238 Maintenance and repair construction
14 311 Food manufacturing
15 312 Beverage and tobacco manufacturing
16 313 Textile mills
17 314 Textile product mills
18 315 Apparel manufacturing
19 316 Leather and allied product manufacturing
20 321 Wood product manufacturing
21 322 Pulp, paper, and paperboard mills and paper product manufacturing
22 323 Printing and related support activities
23 324 Petroleum and coal products manufacturing
24 325 Chemical product and preparation manufacturing
25 326 Plastics and rubber products manufacturing
26 327 Nonmetallic mineral product manufacturing
27 331 Iron and steel mills and manufacturing
28 332 Fabricated metal product manufacturing
29 333 Machinery and equipment manufacturing
30 334 Computer, communications and electronic equipment and components manufacturing
31 335 Electric lighting equipment, household appliance and other electric components manufacturing
32 336 Transportation equipment manufacturing
33 337 Furniture and related product manufacturing
34 339 Other miscellaneous manufacturing
35 43-46 Wholesale and retail trade
36 481 Air transportation
37 482 Rail transportation
38 483 Water transportation
39 484 Truck transportation
40 485 Transit and ground passenger transportation
41 486 Pipeline transportation
42 487 Scenic and sightseeing transportation and support activities
43 488 Transportation support activities
NORTH AMERICAN INDUSTRIAL CLASSIFICATION SYSTEM
INPUT-OUTPUT INDUSTRY CLASSIFICATION

29

a) Key Sectors according to Rasmussen coefficients
Matrix (I A)
-1
, known as Leontiefs inverse, allows one to solve the system in order to
find out the level of production of the ith sector required to satisfy the increment in the final
demand of the jth sector, in one unit. This matrix refers, in consequence, to demand.
Meanwhile the inverse matrix (I D)
-1
that has been derived from the distribution matrix is
Table 3 continued
No. NAICS Industry
44 491 Mail services
45 492 Couriers and messengers
46 493 Warehousing and storage
47 511 Newspaper, periodical, book, and software publishers
48 512 Motion picture and sound recording industries
49 515 Radio and television broadcasting, cable networks and program distribution
50 516 Internet publishing and broadcasting
51 517 Telecommunications
52 518 Internet service providers, web search portals, and data processing
53 519 Other information services
54 521 Monetary authorities
55 522 Credit intermediation and related activities
56 523 Securities, commodity contracts, investments, and related activities
57 524 Insurance carriers, pension funds, and related activities
58 531 Real estate
59 532 Non real state goods rental and leasing
60 533 Lessors of nonfinancial intangible assets
61 541 Professional, scientific, and technical services
62 551 Management of companies and enterprises
63 561 All other administrative and support services
64 562 Waste management and remediation services
65 611 Educational services
66 621 Ambulatory health care services
67 622 Hospitals
68 623 Nursing and residential care facilities
69 624 Social assistance
70 711 Performing arts, spectator sports, and related services
71 712 Museums, zoos and parks
72 713 Amusements, gambling, and recreation
73 721 Accommodation
74 722 Food services and drinking places
75 811 Goods repair and maintenance services
76 812 Personal and laundry services
77 813 Civic, religious, social, professional and similar organizations
78 814 Private households
79 931 Federal State and local Government activities
NORTH AMERICAN INDUSTRIAL CLASSIFICATION SYSTEM
INPUT-OUTPUT INDUSTRY CLASSIFICATION

30

referred to supply. From the elements of these two matrices we can get coefficients that
measure the capacity to generate or to absorb, increments in the various industries. For this
we need to consider first the sum of the elements of each raw, z
i
, and of each column z
j
,
which are called absorption effect and dispersion effect, respectively (United Nations,
2000). The coefficients created by Rasmussen (1956) are developed on the base of each one
of these effects and are obtained by calculating the average of each effect in each industry,
and express these averages as ratios with respect to the global effects.
The estimated quotient based on the absorption effect is known as the Absorption
Power Index and it is defined by the formula:

n
i
n
j
ij
n
i
ij
j
z
n
z
n
U
1 1
2
1
.
1
1
(1)

where: U
j
= absorption power index of industry j, n = number of rows or columns in the
matrix, z
ij
= element ij of matrix (I A)
-1
. This index measures in relative terms, the power
of any given industry to dragging along, or pulling, the whole economy, also called
Backward Linkage.
Similarly, with the dispersion effect, the Dispersion Power Index is calculated by an
equation, based on the distribution matrix:

n
i
n
j
ij
n
j
ij
i
z
n
z
n
U
1 1
2
1
.
1
1
(2)

where: U
j
= dispersion power index, n = number of rows or columns in the matrix, z
ij
=
element ij in matrix (I D)
-1
. This index measures in relative terms the impact produced by
one industry over the rest, also called Forward Linkage. The results are shown in Table 4.


Table 4
No. Industry B.L. No. Industry F.L.
1 30 Computer, communications and electronic equipment manuf. 2.1345 24 Chemical product and preparation manufacturing 3.7554
2 31 Electric lighting equipment and household appliance manuf. 1.5614 6 Oil and gas extraction 3.6203
3 32 Transportation equipment manufacturing 1.4964 35 Wholesale and retail trade 3.4202
4 17 Textile product mills 1.4418 30 Computer, communications and electronic equipment manuf. 2.7352
5 34 Other miscellaneous manufacturing 1.3904 23 Petroleum and coal products manufacturing 2.6556
6 16 Textile mills 1.3437 27 Iron and steel mills and manufacturing 2.6410
7 25 Plastics and rubber products manufacturing 1.3430 61 Professional, scientific, and technical services 2.2770
8 28 Fabricated metal product manufacturing 1.3071 63 All other administrative and support services 2.1490
9 18 Apparel manufacturing 1.3051 25 Plastics and rubber products manufacturing 1.6122
10 21 Pulp, paper, and paperboard mills and manufacturing 1.3009 31 Electric lighting equipment and household appliance manuf. 1.5794
11 9 Electric power generation, transmission, and distribution 1.3005 32 Transportation equipment manufacturing 1.5783
12 19 Leather and allied product manufacturing 1.2862 21 Pulp, paper, and paperboard mills and manufacturing 1.5498
13 29 Machinery and equipment manufacturing 1.2844 9 Electric power generation, transmission, and distribution 1.4761
14 22 Printing and related support activities 1.2546 16 Textile mills 1.4190
15 33 Furniture and related product manufacturing 1.2526 58 Real estate 1.3657
16 36 Air transportation 1.2511 39 Truck transportation 1.3381
17 27 Iron and steel mills and manufacturing 1.2472 28 Fabricated metal product manufacturing 1.3178
18 57 Insurance carriers, pension funds, and related activities 1.2071 1 Crop production 1.1710
19 46 Warehousing and storage 1.1767 57 Insurance carriers, pension funds, and related activities 1.1217
20 5 Support activities for agriculture and forestry 1.1691 51 Telecommunications 1.1110
21 14 Food manufacturing 1.1631 55 Credit intermediation and related activities 1.0970
22 24 Chemical product and preparation manufacturing 1.1630 7 Metal ores and nonmetallic mineral mining 1.0877
23 8 Support activities for mining 1.1247 29 Machinery and equipment manufacturing 1.0132
23 SELECTED INDUSTRIES OF THE 2008 I-O MATRIX OF MEXICO
Backward and Forward Linkages Rasmussen Method

32

In Table 4 the results of the calculation of Backward and Forward Linkages Indexes
according the Rasmussen method, for 23 industries of the I-O Matrix of Mexico of 2008,
are shown. These are the industries with the higher Absorption and Dispersion Indexes
respectively. They form two groups the intersection of which is made up of 13 industries
having at the same time high Backward Linkage indexes and high Forward Linkage
indexes. Those are: Electricity, Textile mills, Paper products, Chemical products, Plastics,
Iron and steel mills, Metal products, Machinery and equipment, Computer and electronic
equipment, Electric lighting equipment and household appliances, Transportation
equipment and Insurance carriers.
b) Linkages determined through the extraction method
We also estimate forward and backward inter-industry linkages through the so-called
extraction method originally postulated by Guido Cella (1984) and further developed by
Eric Dietzenbacher (1993, 1997). We also used two matrices with this method: the
absorption matrix and the distribution matrix. This makes possible to evaluate linkages as
impulses produced either by demand or by supply.
The extraction method works as follows: 1) Given the vector of final demands, the
output is calculated for each of the n sectors or industries; 2) Next, one of these n industries
is hypothetically extracted from the economy by deleting its corresponding row and column
from the matrix A of input coefficients; 3) Using the reduced vector of final demands, the
hypothetical output is computed for each of the remaining n 1 sectors; 4) The effect of
extracting this particular industry is then obtained from the differences (summed over the
remaining industries) between the two types of output. The magnitude of the difference is
indicative of the relevance of the hypothetically extracted industry (Dietzenbacher, et al.,
1993, p.3). The results for the first 23 industries with the highest backward and forward
linkages are presented in Table 5, expressed as the absolute value of the extraction impact
of the corresponding industry and its per cent value with respect to the total of 79
industries. Among these the main non-services industries are the following: Oil and gas,
Electricity, New non-residential construction, Food, Petroleum and coal products, Chemical
products, Plastics, Non-metallic mineral products, Iron and steel, Fabricated metal
products, Machinery and equipment, Computer and electronic equipment, Electric
equipment and household appliances, Transportation equipment and Truck Transportation.

33

Table 5
No. Industry Mill Pesos % No. Industry Mill Pesos %
1 35 Wholesale and retail trade 2,078,601 9.24 35 Wholesale and retail trade 2,047,764 9.13
2 32 Transportation equipment manufacturing 1,205,574 5.36 6 Oil and gas extraction 1,906,972 8.50
3 23 Petroleum and coal products manufacturing 1,198,529 5.33 24 Chemical product and preparation manufacturing 1,508,612 6.73
4 14 Food manufacturing 1,189,604 5.29 23 Petroleum and coal products manufacturing 1,427,409 6.36
5 11 New nonresidential construction 1,181,013 5.25 27 Iron and steel mills and manufacturing 912,760 4.07
6 24 Chemical product and preparation manufacturing 1,106,394 4.92 61 Professional, scientific, and technical services 877,789 3.91
7 6 Oil and gas extraction 895,413 3.98 14 Food manufacturing 833,334 3.72
8 30 Computer, communications and electronic equipment manuf. 887,077 3.94 32 Transportation equipment manufacturing 805,419 3.59
9 9 Electric power generation, transmission, and distribution 778,074 3.46 30 Computer, communications and electronic equipment manuf. 779,187 3.47
10 27 Iron and steel mills and manufacturing 769,470 3.42 9 Electric power generation, transmission, and distribution 661,624 2.95
11 61 Professional, scientific, and technical services 746,159 3.32 39 Truck transportation 625,102 2.79
12 39 Truck transportation 681,153 3.03 63 All other administrative and support services 604,108 2.69
13 31 Electric lighting equipment and household appliance manuf. 538,344 2.39 25 Plastics and rubber products manufacturing 600,765 2.68
14 12 New residential construction 538,186 2.39 31 Electric lighting equipment and household appliance manuf. 597,494 2.66
15 25 Plastics and rubber products manufacturing 494,765 2.20 11 New nonresidential construction 588,680 2.62
16 28 Fabricated metal product manufacturing 449,909 2.00 28 Fabricated metal product manufacturing 476,117 2.12
17 63 All other administrative and support services 444,425 1.98 58 Real estate 431,426 1.92
18 26 Nonmetallic mineral product manufacturing 413,102 1.84 1 Crop production 412,586 1.84
19 2 Animal production 400,806 1.78 55 Credit intermediation and related activities 396,982 1.77
20 1 Crop production 393,580 1.75 26 Nonmetallic mineral product manufacturing 363,355 1.62
21 51 Telecommunications 382,499 1.70 51 Telecommunications 336,676 1.50
22 58 Real estate 367,928 1.64 7 Metal ores and nonmetallic mineral mining 329,506 1.47
23 79 Federal State and local Government activities 353,974 1.57 2 Animal production 321,670 1.43
Total selected industries 77.76 Total selected industries 79.56
Forward Linkage Impact Backward Linkage Impact
23 SELECTED INDUSTRIES OF THE 2008 I-O MATRIX OF MEXICO
Backward and Forward Linkages Impacts by Extraction Method

34

2. Main IOM sectors emitting GHG
The identification of the main polluting industries, in this case the main GHG generators, is
crucial for estimating polluting costs, according to the various input-output models
developed in this regard (Leontief, 1970; Aroche, 2000; Lenzen et al., 2004; Munksgaard et
al., 2005).
a) GHG emissions for the IOM sectors
In order to calculate the GHG emissions in CO
2
equivalent, for each industry of the 79 in
the IOM, we began by analysing the information of the National Greenhouse Gas
Inventory produced by the National Ecology Institute (INE). The GHG emissions are
expressed in Giga-grams of carbon dioxide equivalent (Gg CO
2
eq.), that result of summing
up the emissions of six gases: carbon dioxide (CO
2
), methane (CH
2
), nitrous oxide (N
2
O),
hydro fluorocarbon (HFC), per-fluorocarbons (PFC) and sulphur-hexafluoride (SF6). The
first three together represent 99 per cent of total GHG emissions, while the last three, the
remaining one per cent.
Information regarding these GHG emissions is classified in six sources following
the methodology guidelines of the IPCC which are, in order of importance: 1) Energy, 2)
Industrial processes, 3) Chemical solvents, 4) Agriculture, 5) Land use and forestry and, 6)
Waste disposal. Each of these emission sources is called a sector which is, in turn, divided
into categories and sub-categories. The names given to these sectors and sub-sectors do not
correspond to the economic classification terminology and aggregation criteria of any
national income accounts system.
Next, there was a process of matching the IPCC categories to those of the IOM. For
those industries of the IOM not included explicitly as GHG emission generators, we had to
assign values corresponding to their relative participation in the total of GHG emissions. In
that case we use the relative share of each industry in the Gross Output of the same year,
for estimating these GHG emissions. We, thus, obtained 79 estimates of GHG emissions in
each and every one of the industries in the IOM, measured in CO
2
eq. Then we ordered the
list of emissions from the highest to the lowest. The first 26 GHG emissions generator
industries which are shown in Table 6 produced together 98.5 per cent of all GHG in the
economy.


35

b) Coefficients vector of industries GHG emissions
With the information we had so far gathered or estimated, we calculated the vector of
coefficients of GHG emissions by industry, normalizing the emission values with respect to
the 2003 Gross Output corresponding values, according to the equation:

i i i
x E e
(3)
Table 6
No. Industry GHG Emissions % of
Gg CO2 eq total
1 9 Electric power generation, transmission, and distribution 176,124.8 23.7
2 39 Truck transportation 91,189.0 12.3
3 3 Forestry and logging 69,980.2 9.4
4 64 Waste management and remediation services 61,865.8 8.3
5 40 Transit and ground passenger transportation 61,552.9 8.3
6 10 Water, sewage and natural gas distribution 53,220.4 7.2
7 6 Oil and gas extraction 46,672.5 6.3
8 2 Animal production 39,055.3 5.3
9 26 Nonmetallic mineral product manufacturing 34,429.1 4.6
10 27 Iron and steel mills and manufacturing 30,713.2 4.1
11 1 Crop production 11,031.6 1.5
12 79 Federal State and local Government activities 8,108.5 1.1
13 12 New residential construction 7,520.4 1.0
14 36 Air transportation 6,094.3 0.8
15 24 Chemical product and preparation manufacturing 5,857.3 0.8
16 7 Metal ores and nonmetallic mineral mining 5,146.9 0.7
17 23 Petroleum and coal products manufacturing 4,686.2 0.6
18 11 New nonresidential construction 3,174.9 0.4
19 8 Support activities for mining 2,844.6 0.4
20 43 Transportation support activities 2,400.6 0.3
21 37 Rail transportation 2,230.3 0.3
22 51 Telecommunications 1,922.7 0.3
23 38 Water transportation 1,848.5 0.2
24 42 Scenic and sightseeing transportation and support activities 1,457.6 0.2
25 15 Beverage and tobacco manufacturing 1,342.0 0.2
26 14 Food manufacturing 1,183.0 0.2
Total selected industries 731,652.6 98.5
All industries 742,436.2 100.0
GHG EMISSIONS BY SECTOR IN MEXICO 2008
Selected Industries

36

where: e
i
= annual emission coefficient of industry i; E
i
= annual (2003) Giga-grams of
GHG produced by industry i; x
i
= annual gross output of industry i in millions of Pesos; i =
(1, 2, 3, 79).

Table 7 shows the list of industries with GHG emissions coefficients above the
average, which were 15 out of 79.
There are in Tables 6 and 7, eight industries which are high pollutants both in
absolute and in relative terms: Animal production, Forestry and logging, Electricity, Water,
sewage and natural gas distribution, Non-metallic and minerals, Truck transportation,
Transit and ground passenger transportation and Waste management and remediation
services.
3. Key and High Pollutant sectors
We find that the key or strategic sectors of the Mexican economy according to its IOM
matrix are: first in importance those related directly or indirectly to the fossil fuel
combustion sector: Oil and gas; Electricity, Petroleum and coal products, Chemical
products, Plastics. A second group was that related to metals and minerals: Mining, Iron
and Steel, Metallic products and Mineral products. A third one is related to various types of
machinery, equipment and appliances production: Machinery and equipment, Computer
and electronic equipment, Electric lighting equipment and household appliances and
Table 7
No. Industry Emissions Gross Output Intensity
Gg CO2 eq. Mills Pesos coefficient
1 64 Waste management and remediation services 61,866 7,714 8.020
2 3 Forestry and logging 69,980 25,988 2.693
3 10 Water, sewage and natural gas distribution 53,220 35,139 1.515
4 9 Electric power generation, transmission, and distribution 176,125 543,566 0.324
5 42 Scenic and sightseeing transportation and support activities 1,458 5,856 0.249
6 2 Animal production 39,055 254,792 0.153
7 40 Transit and ground passenger transportation 61,553 414,278 0.149
8 39 Truck transportation 91,189 629,376 0.145
9 26 Nonmetallic mineral product manufacturing 34,429 260,699 0.132
10 38 Water transportation 1,848 22,022 0.084
11 36 Air transportation 6,094 76,258 0.080
12 27 Iron and steel mills and manufacturing 30,713 435,093 0.071
13 37 Rail transportation 2,230 34,468 0.065
14 7 Metal ores and nonmetallic mineral mining 5,147 107,720 0.048
15 6 Oil and gas extraction 46,672 1,132,166 0.041
Selected Industries Total 681,581 3,985,133 0.171
All industries 742,436 20,762,760 0.036
MAIN GHG EMISSIONS COEFFICIENTS BY SECTOR IN MEXICO 2008

37

Transportation equipment. Finally there are also two other industries: Non-residential
construction and Truck transportation. We can say that all in all only 15 sectors or
industries can be defined as strategic, out of the 79 industries that made up the economic
structure according to the North American Classification System (NAICS).
With respect to high GHG emissions industries we find that in the same
classification the important groups are about the same as the strategic ones, except for the
machinery, equipment and appliances related industries which are not high pollutants.
Instead we add five industries related to transportation (other than truck which was already
included): ground, rail, water, air and scenic transportation. Also, three primary activities:
Crop production, Animal production and Forestry and Logging. Finally, we must add,
Waste management and Water sewage and natural gas distribution, for a total of twenty 21
high pollutants sectors.
In the intersection of this big two sets there are three subsets. In the first group in
order of importance, there are only three industries highly strategic for their linkage indexes
and highly pollutant for their absolute and relative GHG emissions, at the same time:
Electricity, Iron and Steel and Truck transportation. It seems convenient to include in this
group another industry, Oil and gas, a high GHG producer and with important forward
linkages but less important backward. In a second group there are two industries: Forestry,
one of the highest GHG emitters but with a low level of inter-industrial connection and
Mining, which always appears second in all groups. This latter group of six industries is
what we may define as the strategic-pollutant sector. On a second level of strategic
importance, but with a high level of energy consumption and GHG emissions are five
industries, which are all associated to transportation services: Air transportation, Rail
transportation, Water transportation, Ground passenger transportation and Scenic
transportation. From our analysis, it is this whole set of eleven industries, which is no doubt
the most relevant to deal with for purposes of GHG emissions mitigation.


38

V. ENVIRONMENT INPUT-OUTPUT MODEL FOR MEXICO

1. Environment Input-Output models
Since Leontiefs works on pollution cleaning up issues (Leontief, 1970, 1973), there have
been various I-O models and analytical devices developed from them. Some advances
appear in input-output textbooks, for instance Miller and Blair (1985, 2009) Ten Raa
(2005); others have been I-O models or analytical instruments applied to particular cases of
sectors and/or regions or countries, for example: Duchin, and Lange (1992, 1994), Kratena
and Scheicher (1999), Idenburg and Wilting (2004), Lenzen, Pade and Munksgaard (2004)
Wilting, Faber and Idenburg (2004), Kelly (2006), and Brink and Idenburg (2007).
From a theoretical approach to the Pollution cleaning model, there have also been
some developments, the most mentioned in the literature are: Steenge (1978), Lowe (1979),
Qayum (1991), Arrous (1994), Lager (1998) and Luptacik and Bhm (1999).
Leontiefs model, which may be called standard Environmental Input-Output model
(EIO), is usually described as:

_
I -A
11
-A
12
-oA
21
I -oA
22
_ j
x
1
x
2
[ = j
y
u
[ (4)

where: A
11
= square matrix of conventional input-output coefficients; A
12
= coefficient
matrix of economic inputs per unit level of abatement activities; A
21
= matrix showing
environmental pollution per unit of production by the conventional sectors; A
22
= matrix
showing pollution generated as a by-product of abatement activities; x
1
= vector of
production levels of the conventional sectors; x
2
= levels of abatement activities; y = vector
of final demand for conventional goods; = diagonal matrix with the percentage of the
pollution which has to be eliminated.
All the above mentioned authors have found that this model is characterised by a
number of assumptions in the way it is formulated that cause limitations of various types
when it comes to be applied for policy evaluation with respect to air pollution (Steenge,
1978, p. 482; Lowe, 1979, p.112; Qayum, 1991, p.428; Arrous, 1994, p.106; Lager, 1998,
p.205; Luptacik and Bhm, 1999, p.265; Brink & Idenburg, 2007, p.3).
Leontief's system as represented by equation (4) has become an important
framework for addressing economy-environment relationships. The approach is, however,

39

characterized by a number of assumptions that cause some problems with the
implementation of the model for environmental policy analysis. These have been pointed
out and dealt with in several studies. We discuss three of them that are relevant for the
analysis of environmental policies with respect to GHG emissions.
In the first place pollution is supposed to be eliminated once it is released into the
environment (surface water, atmosphere, etc.). Although this might be the case for certain
types of pollution (like waste for example), in the case of most gaseous substances (like
greenhouse gases and air pollutants), once they are released into the atmosphere it is hardly
possible to eliminate them (Lager, 1998). Instead, pollution has to be reduced at the source
through the use of less polluting alternative production technologies. This can be achieved
through the substitution of conventional production technology by less polluting production
technologies or else by applying add-on abatement technologies to conventional production
technologies. This has two important implications: (i) abatement activities (and their cost
and effect) are directly related to the pollution at the various specific sources, and (ii)
different substitution and add-on technologies will be available for each of the various
sources, which implies that the cost of reduction and the reduction potentials are sector-
specific.
Secondly, in the standard EIO model, it is assumed that the degree of abatement i.e.
the proportions of pollutants eliminated, represented by in (4) are exogenous to the
model. Moreover, the proportional emission reduction is the same for each sector. With
abatement taking place once pollutants are released into the environment this might be
right, because the abatement cost for a unit of pollution are the same, regardless the source
of pollution. The approach implies that the cost of abatement is spread over the sectors
according to their relative contribution to total pollution. In the context of sector-specific
abatement this will not result in an efficient use of scarce resources to reduce environmental
pollution. In fact, it reflects the instrument of environmental policy called command and
control, prescribing the same abatement technology for each sector.
Taking these limitations into account, the authors reformulated the model in a
suitable way to solve the problems found for the analytical purposes they had in mind.


40

2. Pollution abatement and Technological change models
In our review of the literature we found a very limited number of references regarding
applied I-O models specifically relating climate change, mitigation technologies, impacts
and costs evaluation at a national level. Reviewing all of them, we narrowed our criterion
for selection to those coincidental with the purpose of our study, i.e., the analysis of the
introduction of new technologies for GHG emissions reduction on the economy as a whole.
In this particular sense the EIO models applied to the Dutch economy seemed to be the
most adequate. These were two: a dynamic input-output model called DIMITRI and an
optimization of pollution-abatement technologies model.
In DIMITRI a crucial aspect for the dynamics and the introduction of new
technologies are the variables and equations regarding: investment by sector, capital goods
capacity (existing, expected and planned) by sector, depreciation rates by sector, and the
matrix of capital coefficients. The installed technology is a mix of technologies
implemented in previous periods. As a result of depreciation and new investments, the
installed technology in all sectors changes every period. After installing new technologies,
the technological matrix depicts the new installed mix of technologies. The model estimates
the technological matrices for each period. In the price side of this model, the costs are
compiled from the operational costs, the return of capital and a revaluation of the capital
stock. Sectoral prices are accounted for by the model, prices on labour and other value-
added categories are external (Idenburg, 1998; Idenburg and Wilting, 2000, 2004; Wilting,
Faber and Idenburg, 2004).
The Cost-effective pollution-abatement technologies model, developed by B&I
(2007) is not dynamic and it is built with the purpose of studying the effects of the
application of the best GHG abatement technology per sector, choosing one among various
(at least two) alternatives. The selection is based on a total cost analysis implemented in an
optimization I-O model. The technologies considered for election are all add-on
technologies, that is, they do not imply a change in the product or in the production process.
The model is to be applied to a permit scheme which works under free market rules. That is
to say, the authors privilege in their model an environmental policy of GHG reduction
through a permit market system.

41

Our model has important differences with both models, mostly based on restrictions
imposed by the availability of the information regarding the Mexican Input-Output matrix
and National Income Accounts. We have not enough official data about investment, capital
goods capacity and depreciation rates, by sector per year or a matrix of capital coefficients.
All these are required to make the model dynamic in the way DIMITRI is. Neither have we
got various (more than one) pollution-abatement technologies for each sector of the
Mexican economy, not to mention these technologies in a format that makes it possible to
translate them into input-output vectors. These are basic data to build an optimization
model as the one developed by B&I (2007).
Thus, we have from the outset these limitations, although there are also other
differences with these Dutch models that will become apparent in the assumptions and
specification of the Mexican EIO model. The effects that all this has on policy
recommendations are basically that we are not looking for setting the ideal price for a
carbon permit or a carbon tax, but rather for helping the government to induce the right
technology changes in the right sectors for overall GHG emissions reduction.
3. Objectives and assumptions of the Mexican EIO Model
The main purpose of the EIO model for the Mexican economy is to find out under which
conditions will it be possible for the Mexican Government to fulfil its pledge, made as part
of the Copenhagen Accord (den Elzen, et al. 2010), of reducing the countrys GHG
emissions by 30 per cent below 2002 levels by 2020.
The EIO model for Mexico will show how GHG emissions reducing technologies
applied in selected sectors of the economy will reduce overall emissions through their
direct and indirect effects in the economy and it will try to estimate the economic positive
and negative impacts of these measures for costs and demand growth.
The model will assume the application of a set of abatement technologies in the
strategic-pollutant sector and will estimate first its effectiveness in terms of GHG emissions
reduction and second its impacts on growth and on costs of production. An alternative
scenario will be the Business as Usual (BAU) i.e., no technical change tendency of the
economic structure and GHG emissions, taking as external data the macroeconomic
projections for 2020 of local official or international agencies, for Mexico.

42

In a previous study (SEMARNAT, 2009b) and in the present one we used the Input-
Output approach to measure and identify the most polluting sectors of the economy, the
strategic sectors of the economy and the intersection of these two groups. Each of these
studies was carried out with information for different years and they did in fact yield
somehow different results. In the present case the result was a definition of a subset of both
strategic and high GHG emitting industries we called the strategic-pollutant sector.
Together they represented almost 60 per cent of GHG emissions in Mexico in 2008. The
group, made up of six industries that we define as the strategic-pollutant sector, is no doubt
the set of industries most relevant to deal with for purposes of GHG emissions mitigation.
It is in this sector where the change of technology must occur in the direction from a
currently using technology to the best available GHG reduction technology not in use in
Mexico but somewhere else. This technology should not be an end-of-the-pipe abatement
technology, but one that modifies the production process so as to reduce the GHG
emissions inside the process itself.
The decision on which technology is applied in a variety of existing technologies or
the same production process is not taken by profit maximizing agents operating in a free
market, but by the state decision agency.
4. The model and the scenarios
a) Production pollution model
We start with the usual solution to the Input-Output model:
x
t
= (I A
t
)
-1
y
t
(5)
where: x
t
= gross output measured in economic terms, in time t; A
t
= technical coefficients
matrix in time t; y
t
= final demand or Gross Domestic Product (GDP) vector in time t.
We now introduce the equation of GHG pollution by-products:
x
p
t
= x
t
(6)
where: x
p
t
= vector of pollution levels measured in GHG units; = diagonal matrix of GHG
emissions per unit of output x; t = 0, 1, 2....12, being t =0 (2008) the base year and t = 12
(2020) the end year.
By combining (1) and (2), we get:
x
p
t
= (I A
t
)
-1
y
t
(7)
for all t 6. And,

43

x
p
t
=
+
(I A
+
t
)
-1
y
t
(8)
for all t 7. Where: A
+
t
= estimated matrix for t = 7 modified with new technologies in the
industries of the energy sector represented as new vectors in matrix A.
The basic idea is to estimate absolute and relative differences between the variables
x
p
t = 0
and x
p
t = 12
, that is, the pollution levels, between the base year and the end year,
assuming a fixed GDP annual growth rate g
y
= y
t
/y
t-1
(y = gross domestic product) during
the period t = 2 to t = 12, while the rest, that is, t = 0, t = 1, and t = 2 will be actual data,
taken from official sources.
b) Variables determination:
- Observed variables are all t = 0 variables, plus y
t = 1
, and y
t = 2
. These, and the matrices A
and , are all taken from information gathered from official sources or estimated on its
basis.
- The forecasted vector is y
t = 3,,12

- The exogenous variable are matrix A
+
which is matrix A, modified in the input vectors of
the energy sectors industries, and
+
.
- The unknown vectors, estimated by the model are: x
p
t 0
, x
t 0
; and all p
t 0
.
c) The scenarios
We have four alternative scenarios:
- Scenario (S1): no technical change and a constant growth rate of GDP of 3.5 per cent a
year, from year 2 to year 12. This assumption is founded on the base line of SEMARNAT
study (SEMARNAT, 2009a). We find out how much GHG emissions grow by sector in the
period of constant high growth rate and no technical change.
- Scenario (S2): no technical change and a constant growth rate of GDP of 1.7 per cent a
year, from year 2 to year 12. We find out how much GHG emissions grow by sector in the
period of constant low growth rate and no technical change.
- Scenario (S3): there is a change in technology in six industries of the strategic and high
pollutant group. The technical change introduced is aimed to reduce GHG emissions to the
least possible in all these industries but directly in none other. Technical change is assumed
to initiate in year 4 (2012) but it finds its maturity in year 7 (2015) so from then on matrix
A becomes A
+
. This assumption is based on the observed period between the initial year
that GHG emissions reduction expenses were made in Canada in each sector and the year

44

that effective reductions occurred. It is assumed a constant growth rate of GDP of 3.5 per
cent a year from year 2 to year 12. We find out how much GHG emissions grow by sector
in the period, t = 7 to t = 12 of a constant high growth rate and with technical change.
- Scenario (S4): as in (S3) there is a change in technology in the energy sector, but as in
(S2) it is assumed a constant growth rate of GDP of 1.75 per cent a year from year 2 to year
12. We find out how much GHG emissions grow by sector in the period, t = 7 to 12 of a
constant low growth rate and with technical change.
d) Cost of Pollution model
The model to be used for the calculation of costs is based on the distribution matrix, also
known as the Ghosh matrix (see Dietzenbacher, 1997). This model allows us to determine
the percentage changes in their prices resulting from the effects of a negative externality as
pollution. The individual cost of this bad (as opposed to a good) is estimated as the per-cent
difference between, the vector of gross output, measured in price values x
p
and an estimated
vector x
p
*
we obtained by including the costs of GHG emissions as if they were inputs.
With this model we estimate pollution costs only in Scenario 1, describe above and
considering only data estimated for year 2015 in all the variables.
We start from the vector of gross output in price terms determined by equation:
x
p
= va (I D)
-1
(9)
where: x
p
= gross output vector in price terms; va = value added vector; D = distribution
matrix derived from IOM.
The costs impact equation is defined as:
D
*
= (I + ) D (10)
where: D
*
= increased D matrix; = diagonal matrix with GHG emission coefficient vector
e, in the main diagonal.
The estimation of pollution costs is given by equation:
x
p
*
= va (I D
*
)
-1
(11)
where: x
p
*
= estimated gross output vector (increased by GHG emissions costs) in price
terms.
Pollution relative costs are then determined by:
c
p
= (x
p
*
x
p
) (x
p
)
-1
(12)
where: c
p
= relative differences in costs by industry .

45

VI. TECHNOLOGY CHANGES AND GHG EMISSIONS REDUCTION: THE MODEL
RESULTS

1. Base line or Business as Usual 2008 to 2020.
The first two scenarios describe two trajectories of Business as Usual, which means that
the economy is working without any change in technology aimed to reduce GHG
emissions. We also assume no changes in aggregate demand, in its components or in its
structure by sector, in the period 2010 to 2020, mainly because there is no particular
expected or unexpected event that might lead us to assume otherwise. We assume,
however, a constant growth of Gross Domestic Product (GDP) or final demand at an annual
rate of 3.5 per cent, for this period (2010-2020) in the first scenario, while in the second we
assume a rate of growth of GDP of 1.7. This latter rate is about the same as the population
rate of growth, so it is in fact an assumption of no per capita growth of the GDP in the
whole period.
We utilised the data of the IOM of Mexico for the year 2008, first to forecast the
vector of final demand from 2010 to 2020. With these estimates we use equation (5) of the
model in section V.3.b, to estimate, in turn, the Value of Gross Output for the period 2008-
2020.
Then we used the GHG emissions coefficient by sector estimated in section IV.2.b
by equation (3), converted from a vector to a diagonal matrix called to calculate by
equation (6) the pollution levels for each year. The results are shown in Table 8 and Graph 1.
The calculations in Table 8 indicate that without any technical change, the tendency
for GHG emissions is to grow in both BAU scenarios and the only reduction we find is in the
year 2009 which is explained by the negative rate of growth of GDP, experienced by the
Mexican economy due to the effects of the world crisis.


46

Table 8
Year Gg CO eq. Variation Gg CO eq. Variation
% %
2003 562,785.1 562,785.1
2008 742,436.2 5.7 742,436.2 5.7
2009 708,366.9 -4.6 708,366.9 -4.6
2010 745,581.5 5.3 745,581.5 5.3
2011 771,676.8 3.5 758,256.4 1.7
2012 798,685.5 3.5 771,146.7 1.7
2013 826,639.5 3.5 784,256.2 1.7
2014 855,571.9 3.5 797,588.6 1.7
2015 885,516.9 3.5 811,147.6 1.7
2016 916,510.0 3.5 824,937.1 1.7
2017 948,587.8 3.5 838,961.0 1.7
2018 981,788.4 3.5 853,223.3 1.7
2019 1,016,151.0 3.5 867,728.1 1.7
2020 1,051,716.3 3.5 882,479.5 1.7
2020-2003 488,931.2 86.9 319,694.4 56.8
MEXICO ESTIMATED GHG EMISSIONS TRAJECTORIES
Scenario 1 Scenario 2

47

2. Introduction of technological changes and reductions of GHG emissions
We first selected the industries on which we were going to assume the introduction of a
technological change. These were the industries we found being the highest pollutant and the
strongest inter-connected to the rest in a measurable consistent way. There is, however, one
industry, Forestry, that is not strategic for its measurable linkages to others, but it is crucial to
be included for its high level of pollution. All these were called the strategic-pollutant sectors
and they were the following six: Forestry, Oil and gas, Mining, Electricity, Iron and Steel and
Truck transportation.
The kind of technical change we assume to take place in these industries is a complete
change in the column vector of inputs of each of these industries. That is to say, these vectors
were extracted from the A (technical coefficient) matrix of the IOM of Mexico and replaced
with different vectors which reflect a most efficient production process, that is less pollutant.
Since we could not get anywhere the information required to do this substitution with
new technologies, we took these vectors from the A matrix from the IOM of Canada, to
construct the A
+
matrix that goes into equations (6) and (11) for estimation.
We have three reasons to use Canadas I-O data as a model to follow: first, the
industry classification system used by Canada for its matrix is quite similar to the one Mexico
is using since the beginning of this decade, for National Income Accounts and for Input-
Output data; second, Canadas economy experienced a substantial GHG reduction in many
industries starting in 2003, with technical changes and a set of policies aimed towards GHG
reductions; third the sectors we are taking from them are those in which the emission
coefficient (GHG per unit of output) is the lowest as compared to Mexicos.
It might be questionable to transfer Canadian industrial technologies of some sectors to
the Mexican economy if we take into account the asymmetries (social, political and in the
degree of development) between the two countries, but there is evidence that this type of
transfers are in practice taking place precisely regarding GHG emissions reducing
technologies (see the study of OECD by Hai, et al., 2009).
Thus, we not only substituted Mexican input vectors by Canadian ones, but substituted
the emissions coefficients of these industries in Mexico by the Canadian ones too, in the
matrix , so it became
+
and it was then used in equation (8) for estimation purposes. The
results for all scenarios are shown in Table 9.

48

As expected, the value of GHG emissions measured in Giga grams of CO
2
equivalent
for the years 2015 to 2020 with technical change are less of what they would be with no
technical change. The big reduction comes in 2015 when the technical change starts operating.
The reduction is greater when GDP is growing at the rate of 1.7 (S4) than when it is growing
at 3.5 (S3). But total GHG emissions never go down to 2008 levels, not to mention 2003. All
this can be seen more clearly in Graph 2, there is a clear move down of the tendency line in
both scenarios 1 and 2, resulting from the change in technology in only six industries.

Table 9
Year Scenario 1 Scenario 3 % var Scenario 2 Scenario 4 % var
2008 742,436.2 742,436.2 742,436.2 742,436.2
2009 708,366.9 708,366.9 -4.6 708,366.9 708,366.9 -4.6
2010 745,581.5 745,581.5 5.3 745,581.5 745,581.5 5.3
2011 771,676.8 771,676.8 3.5 758,256.4 758,256.4 1.7
2012 798,685.5 798,685.5 3.5 771,146.7 771,146.7 1.7
2013 826,639.5 826,639.5 3.5 784,256.2 784,256.2 1.7
2014 855,571.9 855,571.9 3.5 797,588.6 797,588.6 1.7
2015 885,516.9 826,139.0 -3.4 811,147.6 756,756.5 -5.1
2016 916,510.0 855,053.9 3.5 824,937.1 769,621.3 1.7
2017 948,587.8 884,980.8 3.5 838,961.0 782,704.9 1.7
2018 981,788.4 915,955.1 3.5 853,223.3 796,010.9 1.7
2019 1,016,151.0 948,013.5 3.5 867,728.1 809,543.1 1.7
2020 1,051,716.3 981,194.0 3.5 882,479.5 823,305.3 1.7
2008-2020 309,280.1 238,757.8 32.2 140,043.3 80,869.1 10.9
MEXICO ESTIMATED GHG EMISSIONS WITH TECHNICAL CHANGE
Gg CO2 eq.

49

3. Changes in GHG emissions by sector
In Table 10 the estimations of GHG emissions by sector are shown for the years 2008, 2015
and 2020 under the Scenarios 3 and 4, for a selected number of industries which were
identified as strategic and/or high pollutants in the 2008 IOM of Mexico. These industries
represented 88 per cent of total emissions in 2008 and their share was reduced in 2020 under
Scenario 3, by 2 points, because they didnt increase their GHG emissions at the same rate as
all the industries of the economy. In fact there are some that reduce in absolute terms the value
of their GHG emissions for 2020, these are: Oil and Gas, Iron and Steel, Forestry and, Mining.
In all these we are introducing a technical change for GHG reduction. The other two that were
considered for a technological change did not in fact reduce their absolute GHG emissions,
Electricity and Truck transportation. The rest of the industries here selected, except maybe
one, were considered as belonging directly or indirectly to the energy sector, for which we
thought a technological change was necessary for increasing their efficiency and reducing
GHG emissions, but we were unable to obtain the information about new available
technologies for these sectors in a format suitable to be used in an Input-Output model.


50

Table 10
No. Industry Base year
2008 2015 2020 Abs. Var. Var. % 2015 2020 Abs. Var. Var. %
9 Electric power generation, transmission, and dist. 176,124.8 183,765.5 218,255.7 42,130.9 23.9 168,332.1 183,135.2 7,010.3 4.0
39 Truck transportation 91,189.0 100,720.8 119,624.7 28,435.7 31.2 92,261.8 100,375.3 9,186.3 10.1
3 Forestry and logging 69,980.2 41,567.8 49,369.5 -20,610.8 -29.5 38,076.7 41,425.2 -28,555.0 -40.8
64 Waste management and remediation services 61,865.8 128,508.0 152,627.2 90,761.4 146.7 117,715.4 128,067.2 66,201.4 107.0
40 Transit and ground passenger transportation 61,552.9 74,129.8 88,043.0 26,490.0 43.0 67,904.1 73,875.5 12,322.6 20.0
10 Water, sewage and natural gas distribution 53,220.4 71,710.1 85,169.1 31,948.7 60.0 65,687.6 71,464.1 18,243.7 34.3
6 Oil and gas extraction 46,672.5 20,278.3 24,084.2 -22,588.2 -48.4 18,575.2 20,208.7 -26,463.8 -56.7
26 Nonmetallic mineral product manufacturing 34,429.1 40,267.6 47,825.2 13,396.1 38.9 36,885.7 40,129.4 5,700.3 16.6
27 Iron and steel mills and manufacturing 30,713.2 6,724.9 7,987.1 -22,726.1 -74.0 6,160.1 6,701.9 -24,011.3 -78.2
36 Air transportation 6,094.3 7,152.6 8,495.1 2,400.8 39.4 6,551.9 7,128.1 1,033.8 17.0
24 Chemical product and preparation manuf. 5,857.3 6,606.2 7,846.0 1,988.7 34.0 6,051.3 6,583.5 726.2 12.4
7 Metal ores and nonmetallic mineral mining 5,146.9 1,670.7 1,984.3 -3,162.6 -61.4 1,530.4 1,665.0 -3,481.9 -67.7
23 Petroleum and coal products manufacturing 4,686.2 5,509.6 6,543.6 1,857.4 39.6 5,046.9 5,490.7 804.4 17.2
37 Rail transportation 2,230.3 3,650.2 4,335.3 2,105.0 94.4 3,343.7 3,637.7 1,407.4 63.1
38 Water transportation 1,848.5 2,477.6 2,942.6 1,094.1 59.2 2,269.5 2,469.1 620.6 33.6
42 Scenic and sightseeing transportation 1,457.6 19,677.3 23,370.4 21,912.8 1,503.4 18,024.7 19,609.8 18,152.2 1,245.4
Selected Industries total 653,069.1 714,416.8 848,503.1 195,433.9 29.9 654,417.2 711,966.3 58,897.2 9.0
All Industries 742,436.2 826,139.0 981,194.0 238,757.8 32.2 756,756.5 823,305.3 80,869.1 10.9
Scenario 3 Scenario 4
ESTIMATES OF GHG EMISSIONS VARIATIONS 2008-2020 BY SECTOR IN SELECTED INDUSTRIES
Gg CO2 eq.

51

4. Technical change and output growth
Table 11 and Graph 3 show the estimated trajectories of Total Gross Output per year in
Millions of Pesos, in the four scenarios simulated. Even though this indicator may not totally
adequate to measure economic growth, it reflects the favourable impulse that a technical
change (introduced through a modification in the technical coefficient matrix) produces in the
general tendency of production.

We can see clearly that technical change, interpreted here as a modification in the Input-
Output matrix, when introduced in some strategic sectors, tends to produce important and
beneficial effects like pollution reduction and output growth, simultaneously.

Table 11
Year Scenario 1 Scenario 2 Scenario 3 Scenario 4
2008 20,762,760 20,762,760
2009 19,288,541 19,288,541
2010 20,553,421 20,553,421
2011 21,272,790 20,902,829
2012 22,017,338 21,258,177
2013 22,787,945 21,619,566
2014 23,585,523 21,987,098 23,585,523 21,987,098
2015 24,411,016 22,360,879 25,446,104 23,309,036
2016 25,265,402 22,741,014 26,336,717 23,705,289
2017 26,149,691 23,127,611 27,258,502 24,108,279
2018 27,064,930 23,520,781 28,212,550 24,518,120
2019 28,012,203 23,920,634 29,199,989 24,934,928
2020 28,992,630 24,327,285 30,221,989 25,358,822
ESTIMATED GROSS OUTPUT 2008-2020
Millions of Pesos

52


53

5. The Cost of Pollution model
We calculated the equations (9) to (12), to estimate Gross Output values for the year 2015
with and without imputed pollution costs, assuming a GDP annual growth rate of 3.5 but using
the Price Model. Table 12 shows the resulting estimations for the total of 79 industries and for
each of 28 selected industries.
Table 12 shows the cost that GHG emissions cause. This represents in general 3.7 of
total Gross Output. This increment in the value of Gross Output is purely caused by GHG
emissions. The 28 selected industries concentrate 74 per cent of this change and as a group
they increased their Gross Output in 5 per cent for pollution imputed costs. These 28 industries
are presented in order of importance according to the size of their estimated pollution costs.
All the industries we defined as either strategic or high pollutants are included in this selected
group of 28.

Table 12
No. Industry GO w/o GO with % of
pollution costs pollution costs absolute relative total D
All industries 24,362,871.5 25,265,340.4 902,468.9 3.7 100
Selected indutries 13,211,216.1 13,875,690.3 664,474.2 5.0 73.6
1 14 Food manufacturing 1,455,287.5 1,542,253.8 86,966.3 6.0 9.6
2 11 New nonresidential construction 1,207,596.5 1,289,114.9 81,518.4 6.8 9.0
3 9 Electric power generation, transmission, and distribution 643,046.4 706,605.5 63,559.1 9.9 7.0
4 20 Wood product manufacturing 76,049.9 133,615.2 57,565.3 75.7 6.4
5 32 Transportation equipment manufacturing 1,091,494.6 1,137,716.9 46,222.3 4.2 5.1
6 27 Iron and steel mills and manufacturing 502,877.0 543,776.9 40,899.9 8.1 4.5
7 30 Computer, communications and electronic equipment manuf. 659,355.4 699,818.3 40,462.9 6.1 4.5
8 24 Chemical product and preparation manufacturing 754,649.0 792,056.8 37,407.7 5.0 4.1
9 23 Petroleum and coal products manufacturing 741,095.5 770,849.1 29,753.7 4.0 3.3
10 12 New residential construction 691,826.7 719,150.9 27,324.2 3.9 3.0
11 26 Nonmetallic mineral product manufacturing 298,116.8 316,605.9 18,489.1 6.2 2.0
12 31 Electric lighting equipment and household appliance manuf. 289,513.6 305,891.6 16,378.0 5.7 1.8
13 28 Fabricated metal product manufacturing 220,994.0 237,190.5 16,196.4 7.3 1.8
14 39 Truck transportation 747,893.3 762,262.1 14,368.8 1.9 1.6
15 61 Professional, scientific, and technical services 690,536.3 702,035.7 11,499.4 1.7 1.3
16 6 Oil and gas extraction 1,296,289.4 1,307,209.8 10,920.4 0.8 1.2
17 25 Plastics and rubber products manufacturing 198,280.9 208,288.4 10,007.6 5.0 1.1
18 29 Machinery and equipment manufacturing 178,893.5 188,073.0 9,179.5 5.1 1.0
19 33 Furniture and related product manufacturing 67,698.6 76,463.1 8,764.5 12.9 1.0
20 40 Transit and ground passenger transportation 492,065.4 500,775.4 8,710.0 1.8 1.0
21 21 Pulp, paper, and paperboard mills and paper product manuf. 162,098.0 169,942.7 7,844.7 4.8 0.9
22 1 Crop production 383,207.9 390,291.3 7,083.4 1.8 0.8
23 7 Metal ores and nonmetallic mineral mining 134,114.4 138,792.3 4,677.9 3.5 0.5
24 36 Air transportation 87,263.4 90,483.8 3,220.4 3.7 0.4
25 10 Water, sewage and natural gas distribution 42,321.6 45,334.4 3,012.8 7.1 0.3
26 37 Rail transportation 41,599.2 42,544.1 945.0 2.3 0.1
27 38 Water transportation 26,391.3 27,207.4 816.2 3.1 0.1
28 3 Forestry and logging 30,660.3 31,340.5 680.2 2.2 0.1
ESTIMATED GROSS OUTPUT AND POLLUTION COSTS FOR 2015
Millions of 2008 Pesos
Differences

54

VII. SUMMARY, CONCLUSIONS AND POLICY RECOMMENDATIONS

1. Summary and Conclusions
As we established in section V.3: the main purpose of building and developing this EIO
model for the Mexican economy has been to find out under which conditions, would it be
possible for the Mexican Government to fulfil its pledge of reducing the countrys GHG
emissions by 30 per cent below 2002 levels, by 2020.
In this section we summarize the findings of our study aimed to meet that objective,
emphasising their implications for state policy. We also point out the limitations this EIO
model has, their causes and the possibilities for further research with the same approach.
a) Strategic sectors and high pollutant sectors
A very important part of this study is section IV, in which we identify two sets of sectors,
strategic and high pollutants, through the use of the Input-Output matrix of Mexico, the I-O
industrial classification system and Mexicos National Inventory of Greenhouse Gases. All
of these are the most recent and reliable data available. These two sets, in principle
different, are made up of very few industries each, showing that the Mexican economy is
highly concentrated in these areas as it is in many others.
We found that key or strategic industries or sectors were in total only 15 out of the
79 that make up the Input-Output matrix according to the North American Classification
System. With respect to high GHG emissions industries we found in the same classification
that the important group consists of a total of twenty one high pollutants sectors, half of
them belonging to the strategic sector just mentioned. Then, in the intersection of these two
sets we found three groups in order of importance. In the first group we considered four
industries highly strategic for their linkage indexes and highly pollutant for its absolute and
relative GHG emissions: Electricity, Iron and Steel, Truck transportation, and Oil and gas.
In a second group there were two industries: Forestry and Mining. The combination of
these two groups formed a set of industries we define as the strategic-pollutant sector. On a
second level we found five industries associated to transportation services: Air
transportation, Rail transportation, Water transportation, Ground passenger transportation
and Sightseeing transportation. These eleven industries are no doubt the most relevant ones
to deal with for purposes of GHG emissions mitigation.


55

b) GHG reduction technologies for strategic and high pollutant sectors
It seemed clear that the industries just defined required attention in the first place for any
GHG mitigation policy related to economic and productive activities. In our opinion as well
as in IPCCs recommendation the only mitigation policy really effective was technological
change.
We found in this respect two alternatives: end-of-the-pipe GHG abatement
technologies or a technical change in the production process and in both cases the
technology had to be sector-specific to be used in the Input-Output matrix. We decided to
look for a technical change in the sectors productive processes, implying a whole change
in the Input-Output vectors of the affected industries or sectors.
Then there was a difficulty in finding at least two alternative technologies per sector
to make the EIO model an optimizing one for choosing the least cost-effective technology
per industry as in B&I (2007). Their EIO model in particular was made up of only thirty
industries and they benefited from another research carried out to make a list of GHG
abatement end-of-the-pipe technologies per sector in the Netherlands by a colleague of
theirs working in the same institute. Our Input-Output matrix consists of 79 industries and
we wanted to make no aggregation for reducing it. Nor did we have the benefit of counting
on such a list of technologies and besides we were not interested in end-of-the-pipe
technologies, so we decided not to build an optimization model but a standard EIO model.
A third decision we made was how to render the EIO model dynamic so as to
estimate the economys annual performance and its GHG emissions through time. We
would have very much liked to copy the dynamic model called DIMITRI built by Idenburg
and Wilting (2004) for the Netherlands, but there is not enough information available in
Mexico regarding capital stock, capacity utilization, investment and depreciation rates per
sector so as to do such a copy for the Mexican economy. Consequently we introduced a
very simple assumption of GDP annual growth for the period 2010-2020 in two possible
scenarios: 3.5 or 1.7 per cent GDP annual growth rate, ceteris paribus, that is to say with no
change in other variables like prices, or demand structure and, of course, no I-O matrix
change except for the technical change. With those assumptions and restrictions, we built
the model in section V.

56

Then, for estimating price or costs changes produced by GHG emissions abatement,
we built another model. This one based on an I-O model, different from Leontiefs standard
and developed originally by Ghosh (1958) later by Gazon (1975, 1979) and Auray et al.
(1980) and more recently by Dietzenbacher (1997). Its main feature is that instead of using
the technical coefficient matrix, usually called A (with coefficients normalized by columns)
it uses the distribution or dbouch matrix called D, or B (with coefficients normalized by
rows). It also uses the value added vector for estimation of Gross Output, instead of final
demand vector (see Miller and Blair 2009, Chs.2 and 12).
We then faced the question of how to introduce a technical change in the selected
industries we had already chosen. There was a problem of availability of the technological
information required for these sectors and also of the format required for such information.
Since we couldnt get this information in the adequate form, we made use of a previous
comparison made with Canadas economic industrial structure expressed in its IOM and its
GHG emissions per sector (see section III). We could get the most recent I-O matrix of
Canada, which follows the North American Industry Classification System (NAICS) as
Mexicos IOM does, as well as the most recent National Greenhouse Gas Inventory of
Canada. This way we were able to calculate the GHG emissions coefficients vector per unit
of output (expressed in the same currency for the same year) and compare it to Mexicos.
We then substituted in our GHG emissions coefficients vector, those of Canadas in the
selected industries. We did the same substitution with the Input-Output column vectors of
the same selected industries in matrix A to simulate a change of technology.
c) Results from the model
The models results are very enlightening about the way GHG emissions are generated (as
by products) in the actual economic structure of the Mexican economy. They show that the
industries related to energy production and energy use as a main input are both highly
pollutant by themselves and highly connected to the rest of the economy so as to expand
this bad influence, generating GHG emissions indirectly.
It is clear that unless there is a big technological change in the energy sector and in
those industries related to transportation services, that is, to the vehicles using oil fuels,
there is little chance to meet the objective of reducing GHG emissions levels, even their
rates of growth. This conclusion also affects one of the most important non-energy related

57

industry which is Forestry. The experience of Canada (see section III) shows that this
industry is strategic in Environmental economics, since it can change from being GHG big
generator to being CO
2
capturer, and reacts very rapidly to a technological change.
We have seen that a technological change aimed to reduce GHG emissions, do reduce
the emissions when this change is applied in strategic and high pollutant sectors. But if the
number of industries having this technical change is increased so as to cover the basic group of
eleven industries we found the most important ones for policy attention, then we could have a
much higher impact in reducing GHG emissions.
In turn, the price model estimated the value of the Gross Output in price terms and
produced an estimate of this GO increased by a set of imputed costs for GHG emissions (as
negative externalities). The differences between these two measurements varied among
industries but, in the average, amounts to 4 per cent of Gross Output.
2. Lessons from the policy debates
The measures originally suggested by IPCC for adaptation and mitigation of Climate Change
were grouped into five categories: market based programs; regulatory measures; voluntary
agreements; scientific research and development (R&D); and infrastructural measures.
Five years later, the 2001 IPCC Report on Mitigation pointed out that, important
considerations in the analysis of climate change mitigation options are, differences in the
distribution of technological, natural and financial resources among and within nations and
regions, and between generations, as well as differences in mitigation costs. And it said
there is also an important issue of equity, namely the extent to which the impacts of climate
change or mitigation policies create or exacerbate inequities both within and across nations
and regions.
With respect to mitigation policies the report recommended that national responses to
climate change can be more effective if deployed as a portfolio of policy instruments to limit
or reduce greenhouse gas emissions[which]... may include emissions-carbon-energy taxes,
tradable or non-tradable permits, provision and/or removal of subsidies, deposit-refund
systems, technology or performance standards, energy mix requirements, product bans,
voluntary agreements, government spending and investment, and support for research and
development. Some of the reports findings on this matter are that Energy efficiency
standards and performance regulations are widely used, and may be effective in many

58

countries, and sometimes precede market based instruments. Voluntary agreements have been
used more frequently, sometimes preceding the introduction of more stringent measures
(IPCC, 2001).
In these reports the need for state intervention arises also from the existence of market
imperfections in each and every economy in the world. The OECD emphasises that putting a
price on GHG emissions, through price mechanisms, has the limitation that they do not
address the full range of market imperfections that prevent emissions to be cut at least cost,
such as information problems (Duval, 2008).
The OECD finds also that empirical analysis indicates that the most important
determinant of innovation in the area of renewable energy technologies is general innovative
capacity. However, it says, public policy makes a difference. Public R&D expenditures on
renewable energies induce innovation, as do targeted measures such as renewable energy
certificates and feed-in tariffs.
Finally, another issue that calls for state action is the issue of equity, namely the
extent to which the impacts of climate change or mitigation policies create or exacerbate
inequities both within and across nations and regions. This implies the need for the
application of state policy measures aiming to prevent or to compensate any inequities that
may result from either climate change impacts or mitigation policies, between sectors or
population groups within a country, and of internationally agreed regulations in the same
direction for inequities between countries.
Despite all recommendations, market oriented policies prevailed in most countries and
they did help, but little, in solving the GHG emissions problem. Nicholas Stern pointed out in
his Review in 2006, after eighteen years of the IPCC foundation, that Climate Change was
the greatest and widest-ranging market failure ever seen (Stern, 2006). Stern also called
for a major change (as opposed to a marginal one) in GHG reductions.
We can interpret what Stern meant by market failure. In the first place this is a
situation in which free market yields an outcome which is not Pareto efficient. The reasons for
that may be, as IPCC and OECD pointed out the existence of market imperfections in most
countries; that we are dealing with a negative externality that cannot be internalized by firms
without government intervention, according to Hepburns opinion; that pollution is a public
good, or rather a public bad, and therefore its price cannot be determined by free market

59

forces; that it is a case of imperfect information as the OECD pointed out (Duval, 2008); or
finally as Hepburn based on Stiglitz considers, that this a market failure due to principal-agent
problems (Hepburn, 2010: Stiglitz, 1991). For all these reasons, there seems to be no doubt
that Climate Change is a great market failure. But there is also another cause of market failure
that applies in this case, the question of equity, emphasized by IPCC and OECD, whether or
not this cause is considered a valid one in modern welfare economics.
Therefore, this great market failure calls for state intervention, even as a second best
policy, that is, with full awareness that it is not possible to have a Pareto efficient solution, in
modern welfare economics terms. And this state intervention has to be as large as the size of
the problem to be solved, in order to produce the major change that is required, as Stern stated.
3. Policy recommendations derived from this study
From what we have learned of the policy debate regarding environmental issues and from
studying Mexicos economic reality and its policy institutions, we think that the federal
government must take immediate and decisive actions for establishing strong regulations in
the use of fossil fuels in industrial processes in order to reduce GHG emissions in specific
sectors we have identified in this study. These regulations must be equal to the highest
standards applied in other countries, and eventually lead to a technological change aimed to
make their productive processes more efficient in terms of the use of energy and of fossil fuel
combustion so as to reduce directly and indirectly GHG emissions. The industries that require
the most attention are essentially in three groups: (a) the energy sector: electricity, mining, oil,
gas, oil products and chemical products; (b) all transportation related industries, from
transportation production, to all services of freight and passengers transportation using fossil
fuels and (c) Forestry and all industries related to it.
Our recommendations, following from the present study, in no way disqualify as a
policy instrument the use of carbon permits, marketable or not, or other instruments that have
been useful in other countries developed or underdeveloped to reduce carbon emissions. Quite
the contrary, the only difficulty we find in this regard is that it is unlikely that a market of
carbon permits will work with the efficacy and efficiency needed by the economy to reduce
GHG emissions. It is indispensable the direct action of the state.


60

a) Technology and Public expenditure
We think that, given the Mexican political and institutional arrangement today, the instruments
to be used for these purposes should be state instruments like taxes, legal pecuniary sanctions,
subsidies, etc. But they must include also public investment since the energy producing firms
are state owned. They desperately need equipment and plants renewal and maintenance. The
state must find ways, directly or through the OECD to buy and adapt GHG reducing
technologies for these sectors, which may be already in operation in other countries. The
required financing for these projects may come also from international sources such as the
World Bank.
Public expenditures are needed as well for Scientific Research and Development
(R&D) in the energy sector to induce technological innovations or adaptations of existing
efficient technologies. Technological change is not only a desired objective but a necessity of
an economy engaged in making more efficient its productive processes in general and in
particular to save energy and protect the environment while at the same time reaching an
adequate rate of growth, that is, sustainable growth.
b) Reforestation
Besides energy, the other one very important sector needing attention in Mexico is Forestry. A
serious policy of reforestation must be implemented, utilising whatever resources the state has
at hand. Forestry is a strategic sector of the economy, since it is a high GHG emissions
generator but can be transformed into a GHG capturer as in the Canadian experience.
4. Further research
In the first place the price model needs to be fully developed so as to calculate the impact on
sectors prices caused by technical change. Secondly, it would be interesting to make a
comparison with some General Equilibrium model developed for Mexico. Thinking this way
it could be possible to extend our own I-O analysis building a Social Account Matrix for
Mexico so as to evaluate social impacts of GHG emissions, on the one hand, and the structure
of demand on the other, and changes in demand induced by technical change.
Another important issue is to find a way to use the results of this I-O model to feed a
Macro econometric model.
But may be the most important point here is to find ways to incorporate new
technologies in all strategic and high pollutant sectors of the Mexican economy in order to

61

simulate a strong technical change in production and distribution activities to find out in a
more realistic way whether it is possible for Mexico to grow and at the same time reduce GHG
emissions and to locate the bottlenecks. The study made by the Instituto Mexicano de la
Competitividad (IMCO) on technological change, might be a good starting point if we can
translate properly each of the technologies the study proposes to use, into the Input-Output
coefficients format. There are other sources that could help to fill the gap, although it may be
expensive to get the information required from them.
Finally, having enough time we even may find useful to build some sort of fixed
capital matrix of the Mexican economy in order to produce forecasts of GHG emissions and
GDP growth for 2020 under various scenarios, based on different new investment and
capacity utilization assumptions including of course technical change in key sectors. We know
that the information for this is not available but we could work something out to build a proxy.


62

BIBLIOGRAPHY

Aroche, F., 2000, Reformas estructurales y composicin de las emisiones contaminantes
industriales. Resultados para Mxico, Nm. 24, Serie Medio Ambiente y
Desarrollo, Santiago de Chile, CEPAL, N.U..

Arrous, J., 1994, The Leontief Pollution model: A systematic formulation, Economic
Systems Research, 6 pp.105-107

Arrow, K., 2007, The case for mitigating greenhouse gas emissions, Real-World
Economics Review, no. 45

Atkinson, A.B., 1970, On the measurement of inequality, Journal of Economic Theory,
No. 2, pp.244-263

Augusztinovics, M., 1970, Methods of International and Inter-temporal Comparison of
Structure, In Anne P. Carter and A. Brdy (eds.) Contributions to Input-Output
Analysis, Proceedings of the Fourth International Conference on Input-Output
Techniques, Vol. I, Geneva, 1968, North-Holland, Amsterdam, pp.249-269

Auray, J. P., Duru, G., y Mougeot, M., 1980, Influence par les prix et influence par les
quantits dans un modle input-output, conomie Aplique, XXXIII, N 3-4, pp.
695-725.

Barker, T., 2008, The Economics of Avoiding Dangerous Climate Change. Editorial Essay
on the Stern Review, Climatic Change, Springer Science and Business Media.

Barker, T., 2004, Economic theory and the transition to sustainability:a comparison of
general-equilibrium and space-time-economics approaches Tyndall Centre
Working Paper No. 62, November, University of Cambridge, Cambridge, UK.

Barker, T., 1998, Use of energy-environment-economy models to inform greenhouse gas
mitigation policy, Impact Assessment and Project Appraisal, vol.16, no. 2, June,
pp. 123-131

Barker, T., 1995, Taxing pollution instead of employment: greenhouse gas abatement
through fiscal policy in the UK, Energy and Environment, vol.6, no. 1, pp. 1-28.

Barker, T., M.S. Qureshi and J. Khler, 2006, The Costs of Greenhouse Gas Mitigation
with Induced Technological Change: A Meta-Analysis of Estimates in the
Literature Working Paper 89, Tyndall Centre for Climate Change Mitigation
Research, Department of Land Economy, University of Cambridge.

Barker, T., S., Scrieciu and L.V., Smith, 2008, World economic dynamics and
technological change: projecting interactions between economic output and CO
2

emissions, Tyndall Working Paper 124, October.

63

Barker, T., T. Foxon, J. Khler, D. Anderson, R. Gross, M. Leach and P. Pearson, 2005,
4CMR (University of Cambridge) and ICEPT (Imperial College London)
submission to the Stern Review on the Economics of Climate Change, unpublished
manuscript, December.

Barker, T., A. Anger, O. Dessens, H. Pollitt, H. Rogers, S. Scrieciu, R. Jones and J. Pyle,
2010, Integrated modelling of climate control and air pollution: Methodology and
results from one-way coupling of an energyenvironmenteconomy (E3MG) and
atmospheric chemistry model (p-TOMCAT) in decarbonizing scenarios for Mexico
to 2050, Environmental Science & Policy, vol. 13, no. 8, December, pp. 661-670.

Boyd, R., and M.E. Ibarrarn, 2008, Extreme climate events and adaptation: an
exploratory analysis of drought in Mexico, Environment and Development
Economics No. 14, pp. 371395

Bravo H.M., J. C. Castro y M.A. Gutirrez, 2011, Anlisis econmico robusto para el
desarrollo de estrategias de bajas emisiones para Mxico, un published paper,
Facultad de Economa, UNAM, Mxico.

Brink, C. and A. Idenburg, 2007 Cost-effective pollution-abatement in an input-output
model, presented at the 16
th
International Input-Output Conference, July, Istanbul,
Turkey.
Cella, G., 1984, The Input-Output Measurement of Interindustry Linkages, Oxford
Bulletin of Economics and Statistics, 70, pp. 705-12.

Cumberland, J.H. and B. Stram, 1976 Empirical Application of Input-Output Models to
Environmental Problems in K. Polenske and J.V. Skolka (eds.) Advances in Input-
Output Analysis. Proceedings of the Sixth International Conferences on Input-
Output Techniques, Vienna, April, 1974, Cambridge MA, Ballinger, pp. 365-382.

den Elzen, M., A.F. Hof, A. Mendoza-Beltrn, G. Grassi, M. Roelfsema, B. van Ruijven, J.
van Vilet, D.P. van Vuuren, 2010, The Copenhagen Accord: abatement costs and
carbon prices resulting from the submissions, Science Direct, Elsevier. Publication
on line: www.elsevier.com/locate/envsci

Dejun, O, N. Gmez, D. Pedregal, M.A. Tobarra and J. Zafrilla, 2008, An energy AGE
model. Forecasting energy demand in Spain, Paper presented to the Input Output
Meeting on Managing the Environment, Sevilla, Spain, 9-11 July.

Dietzenbacher, E., 1997, In vindication of the Ghosh model: a reinterpretation as a price
model, Journal of Regional Science, vol. 37 no. 4, pp. 629-651.

Dietzenbacher, E. and van der Linden, J. A. ,1997, Sectoral and Spatial Linkages in the
EC Production Structure, Journal of Regional Science, vol. 37, no.2, pp.235-257.


64

Dietzenbacher E., van der Linden, J. A. and Steenge, A.E., 1993, The Regional Extracion
Method: EC Input-Output Comparisons!, Economic Systems Research, vol.5, no.2,
pp.185-206.

Dixon, P.B. and M.T. Rimmer, 2009, Forecasting with a CGE model: Does it Work?,
General Paper No. G-197 May, Centre of Policy Studies, Monash University,
Australia.

Drejer, Ina Input-Output Based Measures of Interindustry Linkages Revisited - A Survey
and Discussion, Centre for Economic and Business Research, Ministry of
Economic and Business, Copenhagen, Denmark.

Duchin, F. and G-M Lange, 1992, Technological choices and prices, and their
implications for the US economy, Economic Systems Research, vol. 4, no. 1, pp.
53-76

Duchin, F. and G-M Lange, 1994, The future of the environment, ecological economics &
technological change, Oxford University Press, New York.

Duval, R., 2008, A Taxonomy of Instruments to Reduce Greenhouse Gas Emissions and
their Interactions, OECD Economics Department Working Papers, No. 636, OECD
Publishing, http://dx.doi.org/10.1787/236846121450
Foley, D., 2006, Adams Fallacy, Harvard University Press, Cambridge MA

Foley, D., 2007, The economic fundamentals of Global Warming, paper prepared for the
Workshop on the Economics of Global Warming sponsored by the Schwartz Centre
for Economic Policy Analysis, October 12, 2007.

Fullerton, D. and G. Heutely, 2010, Analytical General Equilibrium Effects of Energy
Policy on Output and Factor Prices The B.E. Journal of Economic Analysis &
Policy, vol.10, no. 2 (Symposium) Article 15

Furman, J., M. Porter and S. Stern, 2002, The Determinants of National Innovative
Capacity, Research Policy, No. 31, pp. 899-933.

Gazon, J., 1975, Transmission de linfluence conomique. Une approache structurale,
Collection de lInstitut de Mathmatiques Economiques, no.13, Sirey, Paris.

Gazon, J., 1979, Une nouvelle methodologie lapproache structurale de linfluence
conomique, conomie Applique, XXXII, 2-3, pp. 301-337.

Ghosh, A., 1958, Input-Output Approach in an Allocation System, Economica, 25, 185-
206.

Hai, I., et al., 2009, Climate Policy and Technological Innovation and Transfer: An
Overview of Trends and Recent Empirical Results, OECD Environment Working
Papers, No. 30, OECD. http://dx.doi.org/10.1787/5km33bnggcd0-en

65

Hansen, B., 1966, The Walrasian System, in H. Townsend (ed.), Price Theory, 1971,
Penguin, New York, pp.107-127

Hepburn, C., 2010, Environmental policy, government, and the market Oxford Review of
Economic Policy, Vol 26, No. 2, pp.117136.

Idenburg, A., 1998, Technological Choices and the Eco-efficiency of the Economy: a
dynamic input-output approach paper presented at the 12
th
International Input
Output Conference, New York, 18-22 May 1998.

Idenburg, A. and H. Wilting, 2000, DIMITRI: a Dynamic Input-output Model to study the
Impacts of Technology Related Innovations Paper resented at the 13
th
International
Input-Output Conference, University of Macerata, Italy, August 21-25, 2000.

Idenburg, A. and H. Wilting, 2004, DIMITRI: A Model to Study Policy Issues in relation
to Economy, Technology and Environment, in: van den Bergh, J.C., and M.
Janssen, Economics of Industrial Ecology; Materials, Structural Change, and
Spatial Scales, MIT Press, Cambridge, MA., pp.223-254.

Intergovernmental Panel on Climate Change (IPCC), 1996, Technologies, Policies and
Measures for Mitigating Climate Change, IPCC, Technical Paper I, WMO UNEP,
Geneva, Switzerland.

Intergovernmental Panel on Climate Change (IPCC), 2000, IPCC Special Report Emissions
Scenarios, IPCC Working Group III, IPPC, Geneva, Switzerland.

Intergovernmental Panel on Climate Change (IPCC), 2001, Climate Change 2001:
Mitigation. IPCC, Working Group III. Ch. 7 Costing Methodologies, Cambridge,
UK, Cambridge University Press, pp. 451-498.

Intergovernmental Panel on Climate Change (IPCC), 2007, Contribution of Working
Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel
on Climate Change (AR4), R.K. Pachauri and A. Reisinger (Eds.) IPCC, Geneva,
Switzerland, pp. 104.

Intergovernmental Panel on Climate Change (IPCC), 2011, web link for history:
http://www.ipcc.ch/organization/organization_history.shtml

Jakob, M., G. Luderer, J. Steckel, M. Tavoni and S. Monjon, 2011, Time to act now?
Assessing the costs of delaying climate measures and benefits of early action
Climatic Change, online issue.

Kelly, A., 2006, An Overview of the RAINS Model, Environmental Research Centre
Report, Environmental Protection Agency, Ireland


66

Kratena, K. and S. Scheicher, 1999, Impact of CO
2
Emissions Reductions on the Austrian
Economy, Economics Systems Research, vol.11, no, pp. 245-261.

Laffont, J.J. and D., Martimort, 2002, The Theory of Incentives. The Principal-Agent
model, Princeton University Press, Princeton, NJ

Lager, C., 1998, Prices of Goods and Bads: an application of the Ricardian theory of
differential rent, Economic Systems Research, 10 (3), pp.203-222

Lambert, P.J., D. L. Millimet and D. Slottje, 2003, Inequality aversion and the natural rate
of subjective inequality Journal of Public Economics, 87 pp.10611090

Layard P.R.G. and A.A. Walters, 1978, Microeconomic Theory, McGraw-Hill, New York

Lenzen, M., L-L Pade, and J. Munksgaard, 2004, CO
2
Multipliers in Multi-region Input-
Output Models, Economic Systems Research, vol.16 no. 4, pp. 391-412

Leontief, W., 1970, Environmental repercussions and the economic structure: an input-
output approach, The Review of Economics and Statistics vol.52, no.3, 262-271.

Leontief, W., 1973, National Income, economic structure and environmental externalities
in M. Moss (ed.) The measurement of Economic and Social Performance, Studies in
Income and Wealth, vol. 38, New York, National Bureau of Economic Research.

Lowe, P., 1979, Pricing problems in an input-output approach to environmental
protection, The Review of Economics and Statistics, 61, pp. 110-117

Luptacik, M. and B. Bhm, 1999, A consistent formulation of the Leontief pollution
model Economic Systems Research vol.11 no.3, pp.263-275

Mercado, R., 2003, Empirical Economy wide Modeling in Argentina, Visiting Professors
Working Papers, Lozano-Long Institute of Latin American Studies. The University
of Texas at Austin, February.

Metz, B., O. Davidson, J-W., Martens, S. Van Rooijen and L. Van Wie Mcgrory, (Eds.)
2000, Methodological and Technological Issues in Technology Transfer IPCC,
Cambridge University Press, UK, pp. 432.

Miller, R. and P. Blair, 2009, Input-Output Analysis Foundations and Extensions, 2
nd

Edition, Ch. 10 Environmental Analysis, Cambridge, U.K., Cambridge University
Press, pp. 446-498.

Munksgaard, J., Wier M., Lenzen M., and Dey Ch., 2005, Using Input-Output Analysis to
Measure the Environmental Pressure of Consumption at Different Spatial Levels
Journal of Industrial Ecology , vol. 9, no. 12, pp.169-185


67

Naciones Unidas, 2000, Manual sobre la compilacin y el anlisis de los cuadros de
insumo-producto, Serie Estudios de mtodos, Departamento de Asuntos
Econmicos y Sociales, Nueva York, Naciones Unidas.

Nordhaus, W., 2007, The Challenge of Global Warming: Economic Models and
Environmental Policy, July, Yale University New Haven, Conn.

Nordhaus, W., 2006, The Stern Review On the Economics of Climate Change Working
Paper 12741, National Bureau of Economic Research, Cambridge, MA.

Nordhaus, W., 1994, Managing the Global Commons: The Economics of Climate Change,
MIT Press, Cambridge, MA

Nordhaus, W., 1992, The Dice Model: Background and Structure of a Dynamic
Integrated Climate Economy Model of the Economics of Global Warming, Cowles
Foundation Discussion Paper No. 1009, Yale University, New Haven, Conn.

Nordhaus, W., 1991, An Inter-temporal General Equilibrium Model of Economic Growth
and Climate Change, in Energy and the Environment in the 21st Century, J. W.
Tester, D. O. Wood, and N. A. Ferrari (eds.), MIT Press, Cambridge, MA.

Nordhaus, W., 1979, The Efficient Use of Energy Resources, Yale University Press, New
Haven, CT, USA.

Nordhaus, W. and J. Boyer, 2000, Warming the World: Economic Modeling of Global
Warming, Cambridge: MIT Press.

Nordhaus, W. and Z. Yang, 1996, A Regional Dynamic General-Equilibrium model of
Alternative Climate Change Strategies, The American Economic Review, vo. 86,
no.4, pp.741-765

Nordhaus, W. and G. Yohe, 1983, Future Carbon Dioxide Emissions from Fossil Fuels,
in National Research Council-National Academy of Sciences, Changing Climate,
National Academy Press, Washington, D.C..

Organisation for Economic Co-operation and Development, 1999, Action Against Climate
Change: The Kyoto Protocol and Beyond, Paris, OECD publications.

Otto, V.M. and J., Reilly, 2008, Directed technical change and the adoption of CO
2

abatement technology: The case of CO2 capture and storage Energy Economics,
30, 28792898.

Pasinetti, L., 1973, The Notion of Vertical Integration in Economic Analysis,
Metroeconomica, vol. 25, no. 1, pp.1-29, February.

Qayum, A., 1991, A reformulation of the Leontief pollution model, Economic Systems
Research, 3, pp. 428-430

68

Rasmussen, P., 1956, Studies in Inter-Sectoral Relations, Copenhagen, Einar Harks.

Rose, A., Y. Cao, and G. Oladosu, 2000, Simulating the economic impacts of climate
change in the Mid-Atlantic Region, Climate Research, Vol. 14, pp. 175-183

Ross, M., 2008, Documentation of the Applied Dynamic Analysis of the Global Economy
(ADAGE) Model Working Paper 08, Research Triangle Institute, September.

Rozenberg, J., S. Hallegatte, A. Vogt-Schilb, O. Sassi, C. Guivarch, H.Waisman and J-C
Hourcade, 2010, Climate policies as a hedge against the uncertainty on future oil
supply. A letter, Climatic Change, DOI 10.1007/s10584-010-9868-8

Ruiz-Npoles, P., 2011, Greenhouse Gas Emissions in Mexico, Relative Costs Estimations
and Policy Implications, paper presented at the 19th International Input-Output
Conference, 13-17 June, 2011, Alexandria VA, USA.

Steenge, A. E., 1978, Environmental repercussions and the economic structure: further
comments, The Review of Economics and Statistics, 60, pp. 482-486.

Stiglitz, J.E., 1991, The Invisible Hand and Modern Welfare Economics Working Paper
No. 3641 National Bureau of Economic Research, Cambridge, MA, March.

Ten Raa, T., 2006, The Economics of Input-Output Analysis, Cambridge University Press,
Cambridge, U.K. Ch.11 Environmental input-output economics, pp.139-150.

Sassi, O., R. Crassous, J-C, Hourcade, J-C, V. Gitz, H. Waisman, and C., Guivarch, 2010,
IMACLIM-R: a modelling framework to simulate sustainable development
pathways, International Journal of Global Environmental Issues, Vol. 10, Nos. 1/2,
pp.524.

SEMARNAT, 2009a, The Economics of Climate Change in Mexico Synopsis, L.M.
Galindo (coord.), SEMARNAT, Gobierno Federal, Mxico.

SEMARNAT, 2009b, Distribucin de los costos del cambio climtico entre los sectores de
la economa mexicana. Un enfoque de insumo-producto, P. Ruiz-Npoles (coord.),
SEMARNAT, Gobierno Federal, Mxico.

Stern, N., 2007, The Economics of Climate Change, the Stern Review, Cambridge UK,
Cambridge University Press.

Stern, N., 2006, The Stern Review on the Economics of Climate Change, H.M. Treasury,
London, UK, October.

Stern, N., 1977, The marginal valuation of income in Artis, M.J., Nobay, A.R. (Eds.),
Essays in Economic Analysis, University Press, Cambridge. Cited by Lambert et al.,
2003 Inequality aversion and the natural rate of subjective inequality Journal of
Public Economics, 87 pp.10611090.

69

Victor, P.A., 1972, Pollution: Economy and Environment, Allen & Unwin Ltd., London.

Wilting, H., A. Faber, and A. Idenburg, 2004, Exploring Technology Scenarios with an
Input-Output Model Paper presented at the International Conference on Input-
Output and General Equilibrium: Data, Modelling and Policy Analysis, September
2-4, 2004, Brussels, Belgium.

World Meteorological Organisation (WMO),1986. Report of the International Conference
on the assessment of the role of carbon dioxide and of other greenhouse gases in
climate variations and associated impacts

Zhai, F., T. Lin, and E. Byambadorj, 2009, A General Equilibrium Analysis of the Impact
of Climate Change on Agriculture in the Peoples Republic of China Asian
Development Review, vol. 26, no. 1, pp. 206225

Statististics and Software Sources:

Instituto Nacional de Estadstica, Geografa e Informtica, Sistema de clasificacin
industrial de Amrica del Norte, Mxico, INEGI.

Instituto Nacional de Estadstica, Geografa e Informtica (2008), Actualizacin del Sistema
de Cuentas Nacionales de Mxico, INEGI, Mxico.

Instituto Nacional de Estadstica Geografa e Informtica, Sistema de consulta de la matriz
insumo-producto:
http://www.inegi.gob.mx/est/contenidos/espanol/proyectos/scnm/mip03/default.asp
?s=est&c=14040

Secretara de Energa, Balance Nacional de Energa, Mxico, SENER.

Secretara de Medio Ambiente y Recursos Naturales, Instituto Nacional de Ecologa (INE)
Inventario nacional de emisiones de gases de efecto invernadero, Mxico,
SEMARNAT.

Statistics Canada, 2008, EnrivoStats, Vol. 2 no. 2, Summer 2008.

Statistics Canada: http://www.statcan.gc.ca/start-debut-eng.html

University of Illinois at Urbana-Champaign, Regional Economics Applicattions Laboratory
(REAL) PyIO: Input-Ouput Analysis with Python.

Wolfram Research, Inc., 2008, Mathematica 7.

Low Carbon Development Strategy For Mexico - An Input-Output Analysis PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Low Carbon Development Strategy For Mexico - An Input-Output Analysis PDF

Uploaded by

Copyright:

Available Formats

Low Carbon Development Strategy for

Mxico: An Input-Output Analysis

You might also like