Download International Trade and Shared Environmental Responsibility by

International Trade and Shared Environmental
Responsibility by Sector. An Application to the Spanish
Economy.
Cadarso, María-Ángeles; López, Luis-Antonio*, Gómez, Nuria; Tobarra,
María-Ángeles
Universidad de Castilla-La Mancha
Facultad de Ciencias Económicas y Empresariales, Plaza de la Universidad n. 2,
02071, Albacete (Spain)
Phone +34 967 599 200 Ext. 2382 . Fax +34 902 204 130.
E-mail: [email protected], [email protected], [email protected],
[email protected]
*Corresponding author
Abstract
The aim of this paper is to define a shared responsibility criterion for analysing the
impact of international trade on CO2 emissions from different sectors. With the
approach proposed it is possible for sectors in a country to account for only a part of the
emissions associated with exported goods and imported goods. The agents considered
as responsible for pollution are sectors of activity by rows, direct emissions linked to
production and sectors by columns, direct and indirect emissions linked to consumption
of inputs and the countries that trade with these sectors. The criterion is applied to the
Spanish economy for the period 2000 to 2005, and proves useful for determining what
economic policies may be suitable for mitigating anthropogenic impact on the
environment.
Key words: CO2 emissions, international trade, shared responsibility between producer
and consumer.
1
1. Introduction
The control and reduction of CO2 emissions generated by a country or a geographical
area, and policies to achieve that control, call for the establishment of an adequate
emission assignation criterion (Ferng, 2003, Peters and Hertwich, 2008a, b). In the
Kyoto Protocol emissions are assigned to the country where they are produced (IEA,
2001), following the so-called producer’s responsibility approach. According to this
principle, a country is responsible for the pollution emitted within its territory when
goods, services or energy are produced, independently of whether they are then
consumed inside or outside the country. Following this criterion an emissions target is
set up for each country and signatory countries are urged to establish mechanisms to
meet it. According to this idea, Directive EC/87/2003 sets up a European emissions
trading market intended to meet the EU’s commitment as a signatory of the Kyoto
Protocol. This is a mechanism for efficiently controlling EU emissions up to 2012, in
which emissions are allocated among chosen sectors following a territorial criterion.
However, the producer’s responsibility principle has some drawbacks that can affect the
degree of commitment by countries to achieving their final targets, i.e. a reduction in
pollutant emissions. The growth of international trade, one of the consequences of
globalisation, boosts the transference of pollution through trade flows, since it is
incorporated into exports and imports. Due to the different pollution intensity
incorporated into exports and imports, a country can transfer or absorb foreign pollution
(Antweiler, 1996, Muradian et al., 2002). Countries can take two different approaches
in commiting to Kyoto emission targets: withdrawing highly-pollutant production
processes or moving them to other countries and then importing the resulting
commodities1. This procedure results in an increase in total worldwide pollution when
countries with less-polluting production technologies outsource production to morepolluting countries, especially if the new producer has not signed up to the Kyoto
Protocol.
This problem, carbon leakage through developed countries imports, is
acknowledged by the Kyoto Protocol and has been analysed in depth in previous
literature (Felder and Rutherford, 1993, IPCC, 2001, Paltsev, 2001, Peters and
1
Causes of offshoring, i.e. delocalisation of production to foreign countries, include cost reduction
(mainly wage costs), economies of scale, access to new markets (for final products or raw materials),
higher elasticity (Abraham & Taylor, 1996) and, especially importantly in our analysis, the existence of
less restrictive environmental laws.
2
Hertwich, 2006, 2008a)2. Pollution also increases due to international trade in
commodities, mainly intermediate inputs, that were previously sourced from the same
geographical area (see Cadarso et al. 2010 for an analysis of this effect). 3
In December 2008 the European Commission endorsed a “Climate and Energy
Package” aimed at regulating the market for emissions allowances from 2013 to 2020 in
the European Union. This package also seeks to avoid the impact of that market on
delocalisation of industry and, therefore, on carbon leakage. To that end, it seeks to
modify emission allocation criteria by incorporating emissions associated with
international trade in commodities. Emissions are allocated in regulated sectors under
the territorial principle and the benchmarking mechanism, however when calculations
are made for all EU countries and not for one particular country, the impact of trade
within the EU is specifically considered. However, EU countries are not prevented from
absorbing or transferring pollution via international trade with other countries that have
not explicitly committed to reducing their emissions. For that reason, the Directive
establishes that those industries that by the end of 2009 are considered at risk of
suffering carbon leakage may not auction their emission allowances, as other sectors of
activity are permitted to do, but can receive greater free allocations. Moreover, the
European Commission also proposes to include a system of European pollution
allowances for commodity importers in sectors at risk of carbon leakage. This is a novel
approach to legislation at international level because it includes imports within a
country’s responsibility (it would be a global responsibility criterion, since it includes
emissions associated with a country’s production and consumption).
The principle of consumer responsibility, defined by Munksgaard and Pedersen (2001),
based on Gay and Props (1993), seeks to deal with the problem of emission leakage and
the impact of international trade on pollution. Under this principle, a country is
responsible for pollution associated with its own energy and commodity consumption,
independently of whether commodities are produced within the country or imported. As
a result, consumer responsibility is calculated by taking producer responsibility, adding
2
See, for example, Peters & Herwitch (2008a), who find most Annex B countries to be net CO2
importers.
3
We can also mention a different problem related to the producer criterion which is outside the scope of
this pape: the difficulty of allocating emissions from international bunkers (UNFCCC, 2005, Faber et al.,
2007, Peters and Hertwich, 2008a, b).
3
emissions associated with imports and removing emissions incorporated into exports.
This principle allows industries and consumers to guide economic development and its
impact on environment through their decisions, reducing their ecological footprint when
they demand products (goods, services or energy) that are low in emissions (Muradian
et al., 2002). As a result, the criterion prevents a country from transferring part of its
pollution as a consumer, though it still transfers responsibility when goods are exported.
It should also be mentioned that the criterion of consumer responsibility suffers from
some drawbacks, such as trespassing on the jurisdictional limits of national power when
the importer country is considered as responsible for emissions associated with imports.
The criterion of shared responsibility is a different way of allocating CO2 emissions. It
entails distributing responsibility between producers and consumers, and is thus an
intermediate procedure between the producer and the consumer criteria mentioned
above. Its main advantage is to commit industries, end consumers and/or countries to
reducing emissions related to both production and consumption. Two different lines of
investigation analyse the best way to calculate shared responsibility. One of them
allocates responsibility for pollutant emissions between countries, taking international
trade flows as a reference. The other allocates responsibility for emissions within a
single country among participating agents. The first method was developed by Ferng
(2003) and Peters (2008), and it considers a country to be responsible for a part of the
emissions associated with its exports and also to a part of those associated with its
imports. This requires a parameter to be established that allows responsibility for
emissions associated with trade to be distributed in an intermediate way between the
producer and the consumer. The second procedure was developed in Bastianoni et al.
2004, Gallego and Lenzen (2005), Lenzen et al. (2007) and Rodrigues et al. (2008). It
seeks to distribute responsibility for the total emission in an economy between different
economic agents, final goods consumers and also workers and firms.
In the context described, our main aim is to design a criterion for shared responsibility
that allows us to identify the sectors of activity that are mainly responsible for
4
emissions4 from an economy within its frontiers or abroad, so that its application can
guide economic policies focused on mitigating anthropgenic impact on the environment.
The methodology developed enables a distinction to be drawn between two concepts of
shared responsibility per activity sector: a) shared responsibility per row (in the inputoutput context), which quantifies how international trade affects direct emissions from a
sector when producing, adding part of the direct emissions linked to exports and
imports; b) shared responsibility by columns, which entails a criterion where sectors are
responsible for all the direct and indirect emissions linked to the purchase of inputs that
cover domestic demand and part of the emissions linked to international trade. For both
rows and columns the stakeholders who share responsibility are sectors, as producers
and consumers (intermediate input purchasers), and the countries that trade with those
sectors. This method is empirically applied to the Spanish economy for 2000 and 2005,
broken down into 46 sectors of activity. Since this is the first study to calculate shared
responsibility by sector of activity using real data for an economy, the results are useful
in evaluating the possibilities for applying in practice the allocation criterion defined.
The paper continues as follows: Section 2 reviews the relevant literature and discusses
in depth the advantages and drawbacks of the different emission allocation criteria;
Section 3 describes input-output techniques that allow emissions from different sectors
within an economy to be allocated according to producer and consumer criteria and,
from that point, enable a country’s emission balance to be calculated; Section 4 sets out
the methodology for the shared responsibility measure for an open economy taking into
account its sectoral structure; Section 5 examines the results and, finally, Section 6
provides a summary of our findings and presents our conclusions.
4
The methodology described seeks properly to allocate emissions associated with energy and goods
production from firms and not from households, whose emissions are mainly associated with transport
and heating. For that reason, the responsibility measures proposed do not include the latter concepts.
5
2. International trade and responsibility for CO2 emissions: a review of the
relevant literature
The international literature on input-output methodology for calculating consumer
responsibility includes a novel paper by Munksgaard and Pedersen (2001) for Denmark
and papers by Ahmad and Wyckoff (2003) for OECD countries and Peters and
Hertwich (2006) for the Norwegian economy. For the Spanish case Sánchez-Chóliz and
Duarte (2004) calculate the CO2 emission balance and analyse contamination associated
with different sectors. They show a slight increase at aggregate level for 1995 that hides
major changes in pollution which are observable only at lower disaggregation levels.
Papers with multi-regional scopes can also be found that tackle trade flows between
areas by using input-output tables for import-producing countries, avoiding the
hypothesis of equal technology (for production and pollution) commonly used in singleregion studies. In these papers, it is possible to select the main trade partners for each
country and input-output data for them; a selected country table is also used as a mean
in order to control for minor traders. Studies along these lines include the papers by
Lenzen et al. (2004), which measures CO2 emissions associated with Danish production
and consumption, Peters and Hertwich (2008b), which calculates emissions associated
with production and consumption of three different gasses in Norway (CO2, NOx and
SO2), and Peters and Herwitch (2008a), which focuses on CO2 emissions for Annex B
and non Annex B countries for 2001 using GTAP databases.
Consumer responsibility is calculated using the input-output benchmark, since it
provides information on final and intermediate goods by sector of activity. Input-output
techniques allow consumer responsibility to be calculated correctly by sectors and by
households making use of emission multipliers. In addition, they enable imports to be
included by deducting responsibility linked to imported inputs that are not finally
consumed within the country because they are used to produce goods and services
subsequently exported, and excluding exports. Moreover, in both cases the use of inputoutput tables not only allows disaggregated analysis but also analysis in direct and
indirect terms. The allocation of direct and indirect emissions allows the sectors
6
ultimately responsible for emissions to be identified, since emissions are transferred
from supplier sectors to consumer ones through the consumption of inputs. This
procedure is equivalent to working with vertically integrated sectors (Pasinetti, 1973) 5.
There are not many papers that analyse emission allocation according to the shared
responsibility criterion, and most of them develop a theoretical approach. They can be
classified according to two main research lines: the first, developed by Ferng (2003) and
Peters (2008), focuses on assigning pollutant emissions among countries by taking into
account international trade flows. Ferng (2003) calculates the emissions for for Taiwan
in 1996 based on the consumer-benefit and producer-benefit principles, which require
the net amount of CO2 sequestered by local ecosystems to be subtracted from the
estimated emissions responsible, and subsequently proposes a criterion of shared
responsibility that includes 50% of the previous two. However the calculations are at
macroeconomic level, so the relevance of structural changes for overall responsibility
cannot be analysed. Peters (2008) discusses the virtues and shortcomings of the
different allocation criteria in terms of international trade and develops shared
responsibility equations for unisectoral and multisectoral economies. This paper does
not, however, develop a procedure for allocating international trade emissions or an
empirical application.
The second line of research does not develop a comprehensive analysis of international
trade, since it focuses on distribution between agents. Bastianoni et al. (2004) develops
a simulation exercise that allows the sectoral emissions from a country to be distributed
in such a way that responsibility for emissions is shared between suppliers and
purchasers and problems of double pollution are avoided. However, no comprehensive
distribution criterion is established. Along similar lines, the papers by Gallego and
Lenzen (2005), Lenzen et al. (2007) and Rodrigues et al. (2008) work on the analysis of
emission distribution between economic agents (sectors of activity, producers and
consumers) and, when producer and consumer responsibility criteria are well
established, seek a suitable criterion for distribution that will allow shared responsibility
to be calculated.
5
Gallego and Lenzen (2005) and shared responsibility researchers in general who do not consider
international trade consider that working with vertically integrated sectors leads to the whole burden
being placed on the end consumer, i.e. households. However, when working with direct data,
responsibility is allocated entirely to producers. When working with vertically integrated sectors our
interpretation is therefore different.
7
Our paper contributes to this literature as one of the first attempts to calculate
responsibility for emissions for a country and its sectors with real data following the
shared responsibility criteria defined in the following section.
2.1 Shared responsibility versus producer and consumer responsibility
The consumer criterion has certain advantages over the producer criterion currently
used. The consumer criterion allows emissions linked to international trade and,
therefore carbon leakage, to be accounted for (Ferng, 2003). It also includes other
emissions apart from those covererd by the signatories to the Protocol, widening the
range of potential emission-reduction policies (Peters, 2008). According to this
criterion, countries should make an effort to reduce emissions associated with inputs
and imports of final goods, boosting the transfer of non-polluting technology to
importer and supplier countries. The transfer of non-polluting technology would have a
great impact because the widespread fragmentation and delocalisation of production,
mainly to low-wage countries, has made countries such as China, Korea, Taiwan or
Eastern European Countries into the main input suppliers for developed countries (see
Grossman and Rossi-Hansberg, 2006, for USA and Cadarso et al., 2008, for the Spanish
economy).
However, although the consumer responsibility principle may be considered a fairer
way of allocating responsibility for emissions in post-Kyoto agreements, it also has its
problems. Peters (2008) points, on the one hand, to the more complex calculations and
the uncertainty that results from that complexity, and on the other hand to the
requirement for countries to make decisions about economic activities performed
outside their jurisdiction and therefore beyond their national political power. This
makes it difficult for countries to hold to agreements and decide what policies to apply,
since they have no control over emissions in other countries.
Moreover, the allocation of emissions to the consumer country or sector implies a
weaker producer commitment to emission reduction, especially for developing countries
(Bastianoni et al., 2004). The criterion of responsibility shared between the producer
and the consumer has arisen in the relevant literature as a mechanism for easing some of
these problems.
8
The main advantage of the shared responsibility criterion is its ability to engage
industries, countries and end consumers in the reduction of emissions arising during
both production and consumption. As an example, the Spanish vehicle industry would
be assigned part of its direct emissions, plus part of those generated by the electricity
consumed in the process and also part of those incorporated in imported automobile
components, so that all the sectors involved are encouraged to look for low-emission
suppliers.
The adoption of the shared criterion helps to ease the implementation of the Kyoto
Protocol in 2012 for less developed countries, see Ferng (2003), since it reduces their
burden of responsibility for emissions associated with exports to developed countries. If
policies aimed at stopping climate change are to succeed, a large number of signatory
countries is a paramount requirement, and the shared responsibility criterion would
encourage developing countries to sign up. The Bali Action Plan, resulting from the 13th
UNFCC Parties Conference, held in December 2007, was the first multilateral
agreement to propose emission mitigating actions for developed and developing
countries (Cascón & Hinojo, 2009). The Copenhagen Agreement, signed on 18th
December 2009, also considers the adoption of mitigating actions for those countries
not included in Annex B, and gives a calendar for developing countries to propose the
establishment of emission reduction goals. However, the possibility of applying the
consumer criterion as a way of achieving lower emission levels is not considered.6 In
this sense, the shared responsibility criterion would be less traumatic than a move to
consumer responsibility, since the latter requires a major change in comparison to
producer responsibility in the total amount of emissions allocated to a country (Lenzen
et al., 2007).
6
Cascón and Hinojo (2009) consider other approaches to encourage the setting up of goals for emission
reduction in developing countries: a) allocating emissions according to population, with a maximum
allowance of CO2 emissions per person; b) allocating emissions according to per capita income, where the
higher the income the greater the requirement for emission reduction.
9
3. Calculation procedure for shared producer and consumer responsibility
criteria between countries.
The total emissions associated with domestic production by firms within a country7, or,
in other words, the producer responsibility (PR), is calculated in the input-output
approach according to the following expression:
PR  e  ( I  A d ) 1  y d   (  y r    y x )
(1)
where  e  is a diagonal matrix of emissions by unit produced by each sector of
activity, I is the identity matrix, A d is the technical coefficients matrix, and  y d  is
the diagonal matrix that captures the final demand met by domestic production.
Elements in matrix  e  are obtained by dividing CO2 emissions per activity sector
(E), by its effective production. From that point it is possible to calculate the emission
multiplier that quantifies direct and indirect emissions by domestic final demand
  e  ( I  A d ) 1 . On the other hand, the responsibility associated with producer
pollution can be decomposed into the total emissions associated with exports (shown in
the diagonal matrix  y x  ) and the pollution associated with the rest of final demand,
final consumption and investment (in diagonal matrix  y r  ).
Expression (1) results in an emissions matrix which can be read by rows and by
columns. The sum of the elements in each row, producer responsibility by rows (PR
rows) covers the emissions produced directly by each sector when producing goods and
services, and equals the emissions allocated following the Kyoto Protocol. The sum by
columns, producer responsibility by columns (PR cols), covers direct and indirect
emissions incorporated into inputs used to produce goods and services to meet final
demand, equivalent to sub-systems (Sraffa) or vertically integrated sectors (Pasinetti).
The results of the two sums are very different for each sector of the economy, although
the total for the whole of the emissions is equal to the figure given by Kyoto.
Consumer responsibility (CR), used among others by Munskgaard and Pedersen (2001),
Ahmad and Wyckoff (2003), Sánchez-Chóliz and Duarte (2004) and Peters and
7
That is, not including direct emissions by households and public administration when consuming energy
goods (gas, coal and oil derivatives).
10
Herwitch (2006, 2008a), includes emissions associated with domestic production sold
within a country plus emissions associated with products made in other countries that
are also the responsibility of the country that buys them as final and intermediate goods
not subsequently exported. From this point of view, emissions included in exports and
in imports required to export are not incorporated, since they must be allocated to the
destination country.
Direct and indirect emissions incorporated into imports coming from country c would
be calculated by:

Ecm  ec  ( I  Act ) 1[A mc I - A d

-1
 y d    y cm  ] 
(2)
  ct [ Acm ( I  A d ) 1  y r ]   ct [ Acm ( I  A d ) 1  y x ]   ct  ycm 
To calculate the above expression it is necessary to know the amount of emissions per
monetary unit of production of the country from which imported goods come ( ec ), the
total coefficient matrix for country c ( Act ), the technical coefficient matrix for goods
imported by country c ( Acm ) and the diagonalised matrix of final demand directly
bought from country c (  y mc  ).  ct  ec  ( I  Act ) 1 is the total emissions multiplier
for country c, which quantifies direct and indirect emissions per unit of total final
demand (domestic plus imported). In the single-region model, expression (2) can be
simplified by assuming that the production technology and pollution in all the countries
involved are the same.8 This assumption allows Act to be replaced by At , ec by e and  ct
by  t , and gives as a result a model of full domestic technology assumption where all
the production rounds are considered. If emissions from all the countries from which
goods have been imported are added, the result is an expression that measures total
direct and indirect pollution, associated with all imports by a country ( E m ):

E m  e  ( I  At ) 1[A m I - A d

-1
 y d    y m ]
(3)
8
This assumption, although common in literature (Munksgaard and Pedersen, 2001, Sánchez-Chóliz and
Duarte, 2004, Peters and Hertwich, 2006), requires many restrictions in calculations, since production
technology, and therefore pollution, is expected to differ from country to country. On the other hand, see
Wiedman et al. (2007) for a theoretical review of single-region and multi-region input-output models for
the assessment of environmental impacts of trade and Andrew et al. (2009) for a quantification of the
errors introduced by various approximations of the full Multi-regional input-output, for national carbon
footprint accounting.
11
Finally, the expression for the consumer responsibility (CR) for the whole of the
economy is:




CR   e  ( I  A d ) 1  y r    e  ( I  A t ) 1 [ A m I - A d
r
t
m
d
   y   [ A ( I  A )
1
r
t

-1

 y r    y m ] 
(4)
m
 y ]    y 
The row summation in expression (4) gives information on the consumer responsibility
rows (CR rows), which consider direct emissions from a sector when producing goods
sold within an economy plus emissions related to imports by an economy of goods and
services in that sector. Consumer responsibility by columns (CR cols) calculates direct
and indirect emissions related to domestic and imported inputs consumed by each sector
(plus emissions of imported final goods in that sector).
Other recent papers that analyse the spread of inter-sectoral emissions define consumer
and producer responsibility differently. Gallego and Lenzen (2005) and Lenzen et al.
(2007) do not consider international trade since they work with a closed economy. For
these authors full producer responsibility includes direct emissions from sectors, while
full consumer responsibility includes emissions from vertically integrated sectors, that
is, direct and indirect emissions associated with final demand consumers. Our
contribution to their work in interpretation by columns is two-fold: a) we consider
responsibility by both sectors and countries because we consider international trade; b)
we allocate responsibility for emissions to sectors as consumers (purchasers) instead of
exclusively to end consumers. In our opinion, to reduce the impact of economic activity
on the environment it is more adequate to allocate emissions to the firms that supply
goods and services than to the consumers of those final goods and services
(consumption, public expenditure, investment9 and exports). Firms can control their
own production processes and can therefore adopt strategies more easily than end
consumers, to reduce impacts on the environment. The exception could be a green tax,
which affects both agents in a similar way, since it imposes a price increase that
encourages intermediate and final consumers to reduce demand for more polluting
goods.
9
In any event, emissions generated by investment goods should be treated differently from other final
demand elements, since the former are within the production system. We consider that responsibility for
emissions produced by investment goods must be accounted for by the businesses buying them, spread
over a number of years according to fixed capital consumption.
12
4. Analysis of shared responsibility of a country.
The idea of applying the shared environmental responsibility criterion to the allocation
of emissions is to make each country accountable for all the emissions associated with
the goods that it produces and consumes and for part of the emissions incorporated into
goods consumed but not produced (imports) and produced but not consumed (exports).
Responsibility for emissions associated with international trade is then shared with the
countries with which a country trades. This procedure seeks to avoid the problem of
carbon leakage and to promote the incorporation of a higher number of countries into
post-Kyoto agreements, mainly developing countries as commented above. On the other
hand, shared responsibility also implies a smaller change than the switch from the
producer to the consumer approach, since is an intermediate position. For some
countries or sectors the application of CR may impose a disproportionate burden, while
for others which are mainly exporters or countries producing for others, the reduction of
the burden may lead to less commitment to and respect for abatement policies.
The expression for the shared responsibility criterion (SR) for a country in a singleregion10 model is:
SR  PR  (1   )CR 

  (1   ) [ A

  ( y r    y x )  (1   )   y r   t [ Am ( I  Ad ) 1  y r ]   t  y m  
   y r    y x
6.1
6.2
t
m
( I  Ad ) 1  y r ]   t  y m 
(6)

6.3
A country is responsible for all the emissions associated with the production of goods
consumed domestically (6.1), plus part of those incorporated into exports (6.2), plus
part of those incorporated into imports (intermediate and final for the first and the
second terms in 6.3 respectively) (6.3). The remaining problem is to find the correct
sharing percentage  . The aim of our proposal is two-fold: to establish a sharing
percentage that on the one hand distributes responsibility for export and import
emissions satisfactorily between countries and on the other hand helps to spread
responsibility adequately at sectoral level. The procedure proposed would lead to a
combination of the two current procedures of shared responsibility. One possible
10
Peters (2008) also shows an expression for calculating shared responsibility for a multi-regional model,
but there is no empirical application.
13
solution would be the proposal by Lenzen et al. (1997) that emissions be allocated by
sectors, where the amount of emissions withheld by suppliers depends on added value
divided by net production (total production less intra-industrial consumption)
11
and
emissions transferred to consumers would therefore be (1-  ). A contribution of this
paper is to apply the inter-sectoral emission criterion to the calculation of shared
responsibility in an open economy, taking into account emissions associated with
exports and imports. The shared responsibility expression for a country A that trades
only with B (which could be the rest of the world) would be:

SR A    y r   A   y x  (1   B )  t [ A m ( I  A d ) 1  y r ]   t  y m 
7.1
7.2

(7)
7.3
The parameters included in diagonalised matrices  A and B , are the quotient of value
added to the net product of sectors of activity in countries A and B respectively.
Expression (7.2) quantifies that part of the responsibility for emissions associated with
exports from country A that remain within that country proportional to  A .
Responsibility for the rest of the emissions is allocated to the countries that consume
those exports. Expression (7.3) measures the emissions for which country A is
responsible, resulting from imports purchased from B, proportional to (1  B ) . B is
responsible for part of the emissions generated by industries that produce goods
exported to A, depending on B’s value added on net production ( B ).12 The method
explained makes a country accountable for that part of emissions that are due to its
consumption, either at aggregate level or by sectors.13 On the other hand, the proposed
allocation of shared responsibility does not lead to double accounting, since some
countries’ exports are other countries’ imports.
11
The advantages of using value added divided by net product as a distribution measure, following
Lenzen et al. (2007) are: invariance with respect to disaggregation of the supply chain, aggregation of the
supply chain and gross or net accounting.
12
For our case, because of the application of a single-region model and the use of a similar technology
assumption,  A = B .
13
However, different sectors of activity share only the emissions that each uses when producing goods
and not those associated with the consumption of intermediate goods by each sector. To add these in, it
would be necessary to work with vertically integrated sectors.
14
This method is compatible with different possibilities for emission allocation by sectors.
Expression (7) results in matrices, since the final demand vectors involved are
diagonalised. The next subsection analyses the calculation of shared responsibility by
rows and also by columns for each sector.
4.1 Shared responsibility by rows.
Sectoral allocation by rows implies that sector i of economy A is held responsible for
the emissions that it generates when producing inputs sold to other sectors that cater for
domestic demand (row in 7.1), plus that part of emissions generated when producing
inputs required for exported goods (row in 7.2), plus part of the emissions generated by
sector i in country B when producing the inputs imported by economy A (first term in
row 7.3), plus part of the emissions generated by sector i in country B when producing
the final goods imported by A (second term in row in 7.3). The advantage of the rows
approach is that it distributes between countries and sectors the emissions accounted for
under the Kyoto Protocol under the producer responsibility principle.14 A drawback is
that it holds a sector responsible for part of the emissions associated with imports
produced by the equivalent sector in a foreign country, even though these imports may
be used by other sectors in the economy or devoted to final demand. However, emission
distribution by rows leads to an allocation of responsibility by sectors similar to that
proposed by the EU market for emissions allowances for 2013 to 2020, though
restricted only to intra-community trade. In this market, emissions are allocated under
the territorial principle and the benchmarking mechanism is calculated at sectoral level
for all EU countries. For that reason, if we focus on EU trade, the row total for shared
responsibility emissions equals the figure proposed by the new European legislation.
However, the allocation of allowances per sector of activity for each country differs
depending on the intensity of the emissions from exports and imports for each country.
The main difficulty in the use of allocation by rows is holding countries responsible for
part of the emissions from the same industry in a different country. For that reason, a
feasible option would be for firms in each country to be held responsible for their own
14
When distribution measures are not used, the sum of the rows gives the emissions from each sector
when producing, as given in national statistics, e.g. in the case of the Atmospheric Emissions Satellite
Account for Spain.
15
emissions through the emissions allowance market, as currently happens, and for the
country to be held responsible for the excess of emissions associated with shared
responsibility. When a country’s emissions are higher than those given according to the
producer criterion additional measures are required. The following are suggested:
encouraging non-polluting energies such as wind and solar by establishing bonus
mechanisms; establishing a tax on carbon content, since firms transfer this to
consumers; levying a special tax on kerosene to internalise part of the large quantities of
emissions from aircraft; developing a policy of environmentally efficient transport with
elements such as boosting of public transport; strictly regulating thermal insulation in
buildings to avoid heat loss; obligatory use of solar energy for water heating in
buildings; establishing minimum temperature settings for air conditioners in public and
private buildings; progressive vehicle registration tax depending on how much a vehicle
pollutes; etc. Following a similar line, the public sector could include an environmental
clause in public procurement contracts so that firms with more environmentally efficient
production processes and those that propose less polluting projects are favoured.
4.2 Shared responsibility by columns
Allocating responsibility for emissions by sectors according to columns implies
reallocating emissions according to vertically integrated sectors (VIS), considering
sectors as consumers and as responsible not just for production but also for purchases
from other sectors. In this way, sector i would be responsible for the emissions
generated in the production of its own inputs used, plus emissions by production
processes in other sectors when producing inputs used by i to cater for domestic demand
(column in 7.1), plus part of the emissions from its own inputs and those of other
sectors that are used to produce goods to be exported (column in 7.2), plus part of the
emissions associated with the inputs imported for its production process to cater for
domestic final demand (first term in 7.3), plus part of the emissions incorporated into
final goods imported for each sector (second term in 7.3).15 The advantages of this
sectoral approach by columns are similar to those of the consumer versus producer
criterion. A sector that uses highly pollutant inputs must account for them, which will
15
According to our “emissions by columns” interpretation, households are responsible only for direct
emissions (heating and private transport mainly). However, that part of the emissions included in final
imports is more difficult to allocate to domestic sectors. One possible solution is to distribute the
responsibility between the producing sector (in a foreign country) and domestic households, which
benefit from consumption.
16
encourage sectors to search for suppliers, or substitute inputs, which are more efficient
in environmental terms. Also, a sector will have a strong incentive to reduce emissions
if its demand sectors require it to do so.
It can be difficult to establish an emission reduction policy based on emission allocation
by columns. On the one hand, it compels firms to reduce emissions of the inputs
demanded, which implies that they have some control over their suppliers’ production
processes or can substitute inputs for which there may in fact be few if any substitutes.
On the other hand, the emissions by columns may be very unbalanced in relation to
rows, and that may imply too high a cost for industry.
The concept of shared responsibility by columns would also be useful for defining the
environmental profile of a product using an environmental label based on CO2
emissions (that could be named ecoemissions or CO2 ecolabel) or of a production
process, similar to ISO 14001 and environmental management systems. Firms could
measure direct and indirect CO2 incorporated into the goods and services that they offer
on markets. In this way, firms recognised as less-polluting in their areas of business
could obtain gains in quality and product differentiation and thus increase their market
share. More importantly final consumers would, through their consumption decisions,
deliberately lead the way on a path of environmentally friendly growth. Also,
considering emissions associated with imports when calculating the CO2 ecolabel could
help to solve poverty and environmental degradation problems in less developed
countries, in a similar way to Fair Trade or Environmental Certifications (Azqueta, et
al., 2006).
There remains a third criterion for distributing responsibility between sectors which is
not analysed in depth in this paper and which would imply the calculation of a shared
responsibility between rows and columns for each sector of activity. This is equivalent
to the proposal by Gallego and Lenzen (2005) and Lenzen et al. (2007) and implies
spreading responsibility between the sum of a row (emissions that are accounted for by
industries) and the sum of a column (emissions that are accounted for by families
according to these authors). This distribution could be complementary to the one
17
proposed here, but we reinterpret emissions by column as responsibility of consumer
sectors and expressly include international trade. 16
5. Environmental responsibility for the Spanish economy.
In an empirical analysis, the responsibility for emissions of the Spanish economy is
caluclated according to producer, consumer and shared criteria at sectoral level. The
main data sources are Input-Output Tables and CO2 emissions per sector of activity
provided by the Atmospheric Emissions Satellite Accounts (CSEA), both published by
INE. Results are calculated for 2000 and 2005 at the lowest available disaggregation
level, the 46 sectors of activity common to both sources, 2000 is the base year. This also
means that emissions incorporated into international trade are adequately captured
since, as Su et al. (2010) show, 40 sectors is a sufficient level of disaggregation to
capture them.
Figure 1 shows that the results under shared responsibility for the Spanish economy are
intermediate between those under the producer and consumer responsibility approaches.
However, as the Figure shows, shared responsibility is closer in growth terms to CR due
to the major increase in emissions under CR for the Spanish economy. This is because
of the great weight and growth of imports. For the Spanish economy as a whole, the
change of criterion from PR to CR means an increase in responsibility of 15.9% for
2000 and 40.8% in 2005, while the shared responsibility criterion shows increases of
9.8% and 34.4% respectively.
It is important to evaluate the impact of the consumer and shared responsibility criteria
compared to the producer criterion not at both aggregate and disaggregated levels, when
the objective is to choose a criterion for allocating responsibility. Importer country firms
and consumers find it hard to assume responsibility for emissions related to imported
goods when this significantly increases their burden. In these cases the most important
aspect is not the amount of pollution responsibility according to the consumer or shared
criteria but what it represents in terms of total producer responsibility (which depends
on the trade balance and on the different emission intensities of products and/or
16
This allocation avoids part of the decomposition between emissions allocated to rows and columns, but
does not ensure the elimination of the divergence with the Kyoto criterion. Moreover, it does not help to
solve the problem of allocation of emissions to firms within a sector of goods from their own or other
sectors produced in other countries and used by firms in any sector or by final consumers.
18
countries, although for our case this latter consideration is not taken on board, since
technology is considered to be similar from country to country).
Figure 1. Producer, consumer and shared responsibility for the Spanish economy,
CO2 gigagrams.
400,000
350,000
300,000
250,000
200,000
150,000
100,000
50,000
0
PR
CR
2000
SR
2005
Source: Own work from input-output tables and CSEA.
5.1 Consumer, producer and shared responsibility by rows (sectors)
Producer responsibility by rows gives the direct emissions from a sector within a
territory. Consumer responsibility by rows subtracts from that figure responsibility in
regard to exports in which the sector participates direct or indirectly (shared
responsibility subtracts it only in part) and adds responsibility for emissions associated
with intermediate and final imports for the whole of the economy for goods produced
by the sector (shared responsibilty adds only a part of this figure). Depending on trade
impact, producer responsibility in the different rows can be higher or lower than
consumer or shared responsibility. In 34 of the 46 sectors, consumer and shared
19
responsibility are higher than producer responsibility by rows, due to the higher content
of imports compared to exports.17
The importance of emissions associated with imports resulting from crude oil and
natural gas must be highlighted. Between them they alone explain a large part of the
discrepancy between producer and consumer/shared responsibility for the Spanish
economy (see Figure 2). These emissions are included in the consumer and shared
measures through imports, but they are not included in the producer-based measure. The
strong dependency of the Spanish economy on oil derivatives and its lack of oilfields
generates a dependency on imports for virtually all crude oil and natural gas consumed
in the economy. When emissions associated with crude oil and natural gas are deducted,
the Spanish CR for 2005 is only 9.8% higher than the PR, and SR is only 5.8% higher.
If emissions due to crude oil and natural gas are deducted for 2000 the relationship is
reversed, with PR being higher by 2.1% and SR by 1.1%.
The data for the Spanish economy show a discrepancy between PR by rows and CR by
rows of less than than 10% in 24 out of the 46 sectors for 2005 and of no more than
30% in a further 12 (Table 1A in the Appendix). In the remaining 10 sectors the
difference is greater than 30% and indeed in 5 of them it is greater than 100% when
compared to either CR by rows or SR by rows. In these 10 sectors, changing to the CR
or SR criterion by rows may lead to a survival problem or a major competitive
disadvantage. Only in one of these sectors (Water transport) are CR and SR lower than
PR. Adopting the SR criterion does not solve the whole of the problem due to the
differences between the producer and the consumer criteria. The sectors related to
information and telecommunication technologies generate low emissions per unit
produced (because their emission coefficient is very low), so the repercussion for them
of the allocation of more emissions than are actually produced would be very small. In
other three - Mechanical engineering, Clothing and furs and Metal ore extraction - the
survival problem could appear. These sectors require special measures, while for the
rest of the economy the SR criterion could be adopted without generating survival
problems for firms.
17
Another reason for the discrepancy is the different emissions associated with a product for the use of
different technologies in the countries of import and export, but this element is not analysed in our
application since similar technologies are considered.
20
Figure 2. Producer, consumer and shared responsibility by rows, selected sectors
where PRrows<CRrows in 2005 (CO2 Gigagrams).
Crude oil and natural gas
Elect., gas and steam product.
and distrb.
Coke, refining and nuclear.
Anthracite, coal, lignite and
peat
Metal ore extraction
Metallurgy
Fishing
0
20,000
40,000
PRrows
60,000
CRrows
80,000
100,000
120,000
SRrows
Source: Own work from input-output tables and CSEA.
Another problem of the responsibility allocation system described, based on CR or SR
by rows, is that firms are accountable for emissions associated with goods in the same
sector imported by themselves or by other firms in the economy or families, goods and
services. That may be considered unfair. Another option would be the method
established by the Climate and Energy Package in the EU, which considers the
possibility of incorporating importers of products with a high risk of carbon leakage into
the Community emission allowance trading system. In this case, the criterion applied to
these products would be overall responsibility (including both domestic production and
imports) rather than consumer or shared responsibility. This would not affect the
feasibility of domestic firms that work with these products, and would compensate them
for the potential loss of competitiveness linked to the emissions market. Another way of
avoiding the unfairness problem is to establish a tax on CO2 exports and imports. So far
as it affects prices, the responsibility (burden) would fall on intermediate and
21
investment goods consumers, firms and final goods consumers and families.18
Moreover, consumer and shared responsibility can be distributed by columns instead of
rows, as detailed in the next subsection.
5.2 Producer, consumer and shared responsibility by columns (VIS).
When emissions are recalculated by columns or vertically integrated sectors, the
responsibility for pollution falls on those sectors that directly and indirectly use highpolluting goods rather than only on those that produce them. The difference between
emissions by rows (observable sectors) and by columns (vertically integrated sectors) is
due to the difference between the amount of responsibility attributable to industry when
producing and the amount due to the consumption of inputs. This difference for
domestic or territorial emissions would be reflected on the PR by rows or columns.
When PR by rows and CR by columns are compared, the difference in emissions
reflects two issues: a) responsibility as producer and as consumer for a sector; b) the
impact of international trade in the responsibility discrepancy. CR by columns reflects
the responsibility of a sector and also considers its international emissions as a
consumer (plus final goods imports of the good that characterises a sector). For that
reason aim to analyse first the difference in responsibilities between producer and
consumer sectors from the criterion of responsibility of a country as a producer (that is,
PR comparing rows and columns) and, second, the impact of trade on direct and indirect
industry responsibility, as the difference between PR by columns and CR by columns,
analysing at the same time the changes that would be entailed by considering SR by
columns as the responsibility allocation criterion. Table 2A in the Appendix shows data
for all the sectors.
Figure 3 shows some of the sectors where the change is greatest between allocating
responsibilty for CO2 within the domestic territory to a sector as a producer (PR by
rows) and as consumer (PR by columns). The sectors shown have production processes
that are not too heavily-polluting, but directly or indirectly use highly-polluting
domestic inputs. They are thus considered pollutant purchasers (Sánchez-Chóliz and
Duarte, 2004) that need other sectors to pollute for them to produce. For instance the
responsibility of the Construction sector as a producer is 3,008 Gg CO2, while as an
18
It must be highlighted that the transfer occurs through the matrix columns.
22
input consumer it is responsible for as much as 52,148 Gg. Moreover, a large number of
the sectors where the repercussion of the change in the criterion is most pronounced are
services sectors, since their emissions are low but their indirect emissions (except those
related to inputs consumed) are much higher. It is also important to highlight the najor
emission by columns associated with Food and Beverages, because of the consumption
of goods that come from agriculture and are thus CO2 intensive, and also for Motor
Vehicles. The discrepancy between the two criteria for the sectors mentioned is such
that PR by rows is 5% of PR by columns for Construction, and 30% for Foods and
beverages.
Figure 3. Responsibility by columns and rows in selected sectors where
PRcols>CRrows in 2005 (CO2 Gigagrams).
Source: Own work from input-output tables and CSEA.
There are also sectors that behave in the opposite way, i.e. they their consumer
responsibility decreases in comparison to their producer responsibility (Figure 4). These
sectors are usually intermediate goods and services producers such as Electricity
Production or Distribution, Non-metallic Mineral Products, Water and Terrestrial
transport, Crude Oils, Metallurgy and Agriculture. Their activities as consuming sectors
is much less contaminant than their activities as producers. So for Agriculture PR by
rows is 1.6 times higher than PR by columns and for Non-metallic Mineral Products it
is 4.5 times higher. The strong imbalance between the responsibility of industries as
23
producers and consumers makes it difficult for PR rows to be chosen as the only
criterion for allocating responsibilities. Only in 3 out of the 46 sectors are the
differences smaller than 30% and for most cases they are over 70%.
Figure 4. Responsibility by columns and rows in selected sectors where
PRrows>PRcols in 2005 (CO2 Gigagrams).
Electric energy, gas and steam
production or distribution
Other non-metallic mineral
products
Terrestrial transport
Cokery, refinery and nuclear.
Metallurgy
Agriculture, stockbreeding and
hunting
-10,000
0
10,000
PRrows
20,000
30,000
PRcols
40,000
CRcols
50,000
60,000
70,000
80,000
90,000
100,000
SRcols
Source: Own work from input-output tables and CSEA.
Graphs 3 and 4 also show the impact of trade on the total emissions of a sector, when
comparing total emissions under the producer criterion (PR cols) with total emissions
under the consumer criterion (CR cols) and under SR. In most of the sectors selected
consumer responsibility by columns is higher than producer responsibility by columns,
that is, emissions associated with imports by these sectors are higher than emissions
associated with exports. Once more, SR by columns gives an intermediate measure
between the other two, but one that it is closer to the consumer than to the producer
criterion.
A comparison of the differences in the first three measures in the two graphs shows that
they are much stronger between PR rows and columns than between PR columns and
CR columns. That is, the impact of trade on the change from the producer to the
consumer criterion by columns is smaller than the impact of the change from emissions
from industry as a producer to industry as a consumer. The final result is such that the
24
differences between SR by columns and producer responsibility by rows, or the Kyoto
allocation criterion, are very significant for all sectors. As an example, the emissions
allocated to Construction according to Kyoto are 4.5 of the total emissions allocated by
the SR by columns. Also for Electricity and Gas Production Kyoto emissions are 227%
of the SR criterion. For Metallurgy the figure is 839%.
It is concluded, contrary to the findings for the responsibility analysis for products or
rows, that a criterion of responsibility allocation that takes international trade into
account for Spain and also emissions associated with intermediate inputs is hard to
apply because of the differences in responsibility for firms. Applying the SR criterion
by industries reduces this discrepancy, but only by a little. That does not mean that there
are no suitable policy options. We consider that the main advantage given by the partial
use of SR by columns is that firms would encourage their suppliers to improve their
environmental efficiency, which would justify their adoption. Also, the responsibility
could be shared between input suppliers and consumers and final consumers (as
suggested by Gallego and Lenzen, 2005), making the application of the criterion less
traumatic.
An alternative policy could be to offer a bonus to firms that are more efficient in direct
and indirect terms by establishing a CO2 ecolabel that allows them to differentiate their
products for consumers (intermediate and final), so that they can, through their
decisions, guide the economy to a sustainable development path. Another possibility
would be to set two clearly differentiated emission reduction goals: one for direct
emissions, such as the one now established in the emissions market, and the other for
total emissions calculated from the consumer criterion by columns or the SR one. In the
latter emission reduction goals would be chosen based on the identification of firms that
are directly and indirectly more efficient in environmental terms for each industry, so
that the penalty for firms that do not achieve the goals would not compromise their
survival. As an example, the purchase price of emission allowances must not be higher
than 5-10% of the wage cost.
6. Conclusions
The achievement of the emissions reduction goals proposed for the signatory countries
of the Kyoto Protocol, e.g. for Spain, requires a wide range of policies to be established.
25
Correctly identifying responsibility for pollutant emissions can lead to more effective
policies. The Kyoto Protocol uses a territorial emission allocation criterion, so
responsibility is allocated to the polluter. However, globalisation has led to an increase
in international trade in final and, especially, intermediate goods, which has
repercussions on emission allocation. Through international trade pollution is
transferred between developed Kyoto signatory countries and non-signatories, resulting
in a carbon leakage that may lead to an increase in emissions at global level. Including
international trade in the responsibility for emission allocation procedures would help to
solve the problem. This is what the consumer responsibility criterion achieves.
However, the change in the allocation from the producer to the consumer criterion may
be too sharp. At global level, the results indicate that under the consumer criterion the
Spanish economy’s responsibility for CO2 emissions would have been 40.8% greater in
2005. The SR criterion may help to ease these problems, since it is an intermediate
procedure that involves both consumers and producers. Applying SR in the Spanish
economy would have increased Spain’s responsibility as a producer country by 9.8% in
2000 and by 34.4% in 2005.
Analysis by sectors and columns allows the responsibility for emissions of the country
that produces them to be transferred to its firms. The method proposed allows SR to be
identified by rows, where only direct emissions from production are included, but these
sectors share the emissions linked to international trade in their product.
The
differences observed between PR by rows compared to CR by rows and SR by rows are
only significant for a small number of sectors. Emissions associated with crude oil and
natural gas are the principal reason for difference between PR, CR and SR, all by rows,
for Spain. However, it is hardly fair for firms in a sector to be considered as responsible
for emissions associated with imports by other sectors or by final consumers of a
product. Another option is to follow the guidelines of the Climate and Energy Package,
which proposes that product importers in sectors at risk of carbon leakage should be
included in the European emission allowance system. A third option is the
establishment of a frontier tax on emissions linked to imports, reducing the possibility
of emissions leakage.
In any case, under the SR criterion by rows, the imports included in the European
emissions market and the burden of a tax must be set up in such way that less of a
26
burden is imposed on emissions associated with international trade than those related to
production to meet domestic demand. In this way, Europeans would be responsible for
only part of the emissions associated with international trade, reducing the carbon
leakage problem (compared to the Kyoto allocation criterion), which would give
stronger coverage than the Kyoto signatory countries.
SR by columns is also defined, where sectors are accounted for by emissions linked to
inter-sectoral emissions and part of those included in international trade in the products
consumed by the sector. This eliminates the drawback of responsibility by rows since it
relates the emission responsibility for imports to the goods consumed by the sector. But
it has an important drawback; the difference between PR rows (Kyoto) and SR cols is
greater than 70% for most sectors of activity, which would clearly lead to opposition to
its application. For instance the Clothing sector goes from responsibility by rows as a
producer to the tune of 157 CO2 Gg (Kyoto) to 3,481 Gg under CR by cols and 2,626
Gg. under SR by cols. This could prevent the application of effective emissions
reduction policies in similar cases without tackling income profitability or survival
problems for firms. On the other hand, sectors with a strong exporter or intermediate
goods producer profile may lose interest in or lack commitment to the application of
emission reduction measures when the consumer responsibility criterion reduces the
emissions for which they are accountable since the burden falls mainly on others
(consumers) as mentioned by Peters (2008), and SR does not solve this problem. For
example, the responsibility of Other Non-metallic Mineral Products changes from
50,260 Gg CO2 as producer under the criterion by rows (Kyoto) to 2,162 Gg as a
consumer by columns and 5,871 under shared responsibility by columns. From this
perspective the policy options considered would favour the creation of an ecolabel that
shows consumers which firms are more environmentally friendly or of setting different
goals for reducing direct and indirect emissions so that the survival of firms is not
affected.
27
BIBLIOGRAPHY
Abraham, K. and Taylor, S., 1996. Firm’s use of outside contractors: theory and
evidence. Journal of Labour Economics, 14, 394-424.
Ahmad, N. and Wyckoff, A., 2003. Carbon dioxide emissions embodied in
international trade of goods. OECD Science, Technology and Industry Working Paper,
15.
Andrew, R., Peteres, G.P. and Lennox, J. 2009. Approximation and regional
aggregation in multi-regional input-output analysis for national carbon footprint
accounting, Economic System Research, 21 (3). 311-335.
Antweiler, W., 1996. The pollution terms of trade. Economic System Research,
8, 4, 361-365.
Azqueta, D., Delacámara, G. and Sotelsek, D., 2006. Degradación ambiental,
endeudamiento externo y comercio internacional. Cuadernos Económicos de ICE, 71,
115-132.
Bastianoni, S., Pulselli, F. M. and Tiezzi, E., 2004. The problem of assigning
responsibility for greenhouse gas emissions. Ecological Economics, 49, 253– 257
Cadarso, M.A., Gómez, N.; López, L.A. and Tobarra, M.A., 2008. The EU
enlargement and the impact of outsourcing on industrial employment in Spain, 19932003, Structural Change and Economic Dynamics, 19 (1), 95-108.
Cadarso, M.A., Gómez, N.; López, L.A. and Tobarra, M.A., 2010. CO2
emissions of international freight transport and offshoring: Measurement and allocation,
Ecological Economics, 69, 1682–1694.
Cascón, P.and Hinojo, P. (2009). La dimensión económica internacional en la
lucha contra el cambio climático, Revista de economía ICE, 149-160
European Commission, 2003. Directiva CE/87/2003. European Communities,
Luxembourg.
Faber, J., Boon, B., Berk, M., den Elzen, M., Olivier, J., Lee, D., 2007. Aviation
and maritime transport in a post 2012 climate policy regime. Report 500102 008.
Netherlands Environmental Assessment Agency (MNP, The Netherlands.
Ferng, J. J., 2003. Allocating the responsibility of CO2 over-emissions from the
perspectives of benefit principle and ecological deficit. Ecological Economics, 46, 121141.
28
Felder, S. and Rutherford, T., 1993. Unilateral CO2 reductions and carbon
leakage: the consequences of international trade in basic materials. Journal of
Environmental Economics and Management, 25, 162-176.
Gay, P. W. and Proops, J. L. R., 1993. Carbon-dioxide production by the UK
economy: an input-output assessment. Applied Energy, 44, 113-130.
Gallego, B. and Lenzen, M., 2005. A Consistent Input–Output formulation of
shared Producer and Consumer Responsibility. Economic Systems Research, v. 17, n. 4,
365–391.
Grossman, G.M., Rossi-Hansberg, E., 2006. The rise of offshoring: it's not more
wine for cloth anymore. Symposium The New Economic Geography: Effects and
Policy Implications, Jackson Hole, Wyoming.INE, 2005. Contabilidad nacional de
España. Marco input-output. Serie 1995-2000. Madrid, Instituto Nacional de
Estadística.
INE, 2006. Cuentas Satélite sobre Emisiones Atmosféricas. Serie 1995-2000.
Madrid, Instituto Nacional de Estadística.
International Energy Agency (IEA), 2001. CO2 from fuel combustion, OECD,
Paris.
IPCC, 2001. Climate Change 2001: mitigation. Cambridge University Press,
Cambridge, UK.
Lenzen, M., Murray, J., Sack, F. and Wiedman, T., 2007. Shared producer and
consumer responsibility – Theory and practice. Ecological Economics, 61, 27-42.
Lenzen, M.; Pade, L.L. and Munksgaard, J., 2004. CO2 multipliers in multiregion input-output models. Economic System Research, 16, 4, 391-412.
Munksgaard, J. and Pedersen, K., 2001. CO2 accounts for open economies:
producer or consumer responsibility? Energy Policy, 29, 327-334.
Muradian, R., O’Connor, M. and Martinez-Alier, J., 2002. Embodied pollution
in trade: estimating the ‘environmental load displacement’ of industrialised countries,
Ecological Economics, 41 (1), 51-67.
Paltsev, S., 2001. The Kyoto Protocol: regional and sectoral contributions to the
carbon leakage. The Energy Journal, 22 (4), 53-79.
Pasinetti, P. L., 1973. The notion of vertical integration in economic analysis”,
Metroeconomica, 25,1-29.
Peters, G. P., 2008. From prodroduction-based to consumption-based national
emission inventories. Ecological Economics, 65, 13-23.
29
qPeters, G. P. and Hertwich, E. G., 2006. Pollution embodied in trade:
the Norwegian case. Global Environmental Change, 16, 379-389.
Peters, G. P. and Hertwich, E. G., 2008a. CO2 embodied in international trade
with implications for global climate policy. Environmental Science and Technology, 42
(5), 1401-1407.
Peters, G. P. and Hertwich, E. G., 2008b. Post-Kyoto gas inventories: production
versus consumption. Climatic Change, 86, 51-66.
Rodrigues, J., Domingos, T., 2008. Consumer and producer environmental
responsibility: comparing two approaches. Ecological Economics 66, 533-546.
Sánchez-Chóliz, J. and Duarte, R., 2004. CO2 emissions embodied in
international trade: evidence for Spain. Energy Policy, 32, 199-2005.
Su, B., Huang, H.C., Ang, B.W. and Zhou, P. 2010. Input-output analysis of
CO2 emissions embodied in trade: The effects of sector aggregation, Energy
Economics¸ 32, 166-175.
UNFCCC, 2005. Information on Greenhouse Gas Emissions from International
Aviation and Maritime Transport. FCCC/SBSTA/2005/INF.2. United Nations
Framework Convention on Climate Change.
30
Appendix
Table 1A. Producer, consumer and shared responsibility by rows, Gg of CO2
Year 2000
Year 2005
SRrows/
SRrows/
PRrows CRrowsSRrows
PRrows CRrowsSRrows
PRrow (%)
PRrows (%)
(A)- 01 Agriculture, stockbreeding and hunting
(A)- 02 Silviculture and forestry exploitation
(B)- Fishing
(C)- 10 Anthracite, coal, lignite and peat
(C)- 11 Oil crudes, natural gas and uranium
(C)- 13 Metallic minerals extraction
(C)- 14 Non-Metallic minerals extraction
(D)- 15 Food and beverages
(D)- 17 Textile industry
(D)- 18 Clothing and furs
(D)- 19 Leather and footwear
(D)- 20 Wood & products of wood and cork
(D)- 21 Pulp, paper & paper products
(D)- 22 Printing & publishing
(D)- 23 Mineral oil refining, coke & nuclear fuel
(D)- 24 Chemicals
(D)- 25 Rubber & plastics
(D)- 26 Non-metallic mineral products
(D)- 27 Basic metals
(D)- 28 Fabricated metal products
(D)- 29 Mechanical engineering
(D)- 30 Office machinery
(D)- 31 Electric machinery and materials
(D)- 32 Electronic materials
(D)- 33 Medical, precision and optical instruments
(D)- Motor vehicles, trailers and semi-trailers
(D)- 35 Other transport equipment
(D)- 36 Furniture, miscellaneous manufacturing
(D)- 37 Recycling
(E)- 40 Electricity, gas and water supply
(E)- 41 Water collecting, treatment and distribution
(F)- 45 Construction
(G)- 50-52 Motor vehicles and reparation
(H)- 55 Hotels & catering
(I)- 60 Inland transport
(I)- 61 Water transport
(I)- 62 Air transport
(I)- 63 Supporting and auxiliary transport activities
(I)- 64 Communications
(J)- 65-67 Financial intermediation
(K)- 70-74 Real estate activities and business services
(L)- 75 Public administration
(M)- 80 Education
(N)- 85 Health and social work
(O)- 90-93 Other community, social & personal serv.
(P)- 95 Private households with employed persons
Total
8,296
101
3,143
1,078
361
176
495
5,018
1,423
133
121
445
2,685
198
20,300
6,890
128
45,041
13,353
396
312
7
98
9
7
222
68
223
317
91,048
546
2,654
5,236
2,780
19,477
2,739
7,404
1,553
232
208
567
743
542
915
1,004
0
6,910
98
3,634
3,556
45,193
1,076
404
4,873
1,499
172
96
425
2,423
191
22,245
7,223
115
38,042
13,540
418
401
12
104
18
11
192
62
222
321
95,535
535
2,648
4,787
2,779
16,642
1,361
4,915
1,374
225
199
562
743
548
914
989
0
7,782
101
3,354
2,087
27,409
758
440
4,904
1,474
158
104
432
2,504
194
21,952
7,122
120
40,887
13,484
410
366
11
102
15
9
199
64
222
321
93,454
540
2,651
5,076
2,780
18,219
1,883
5,810
1,454
229
206
565
743
543
915
998
0
93.8
99.7
106.7
193.6
7592.5
431.0
88.8
97.7
103.6
119.1
86.0
97.0
93.2
97.8
108.1
103.4
93.5
90.8
101.0
103.5
117.3
160.7
104.3
171.3
132.9
89.4
94.4
99.6
101.1
102.6
98.9
99.9
96.9
100.0
93.5
68.7
78.5
93.6
98.9
99.0
99.7
100.0
100.1
100.0
99.4
-
8,660
112
2,563
842
295
204
565
6,158
1,750
157
141
529
3,316
224
20,572
8,294
153
50,260
14,205
465
566
8
116
10
8
249
84
242
370
92,445
679
3,008
5,878
3,134
21,921
3,580
7,455
1,755
255
228
624
817
596
1,010
1,134
0
7,654
117
3,625
5,304
82,565
3,265
606
6,135
2,274
258
148
576
3,357
228
25,176
9,257
158
46,486
16,994
554
866
24
135
25
14
249
74
272
439
110,944
680
3,029
5,637
3,156
19,939
2,042
5,469
1,650
261
232
666
817
606
1,011
1,138
0
8,241
112
3,019
2,833
76,376
1,949
591
6,140
2,078
222
146
558
3,341
226
24,662
8,958
156
48,014
16,117
521
742
19
130
20
12
249
76
262
495
102,713
679
3,020
5,786
3,144
20,814
2,546
5,782
1,699
257
229
638
817
597
1,010
1,135
0
248,692 288,231 273,050
109.8
265,637 374,113 357,133
31
95.2
100.3
117.8
336.5
25890.3
955.6
104.5
99.7
118.7
141.4
103.3
105.4
100.8
100.9
119.9
108.0
101.9
95.5
113.5
112.0
131.1
241.4
112.2
202.5
152.0
100.0
91.0
108.1
133.7
111.1
100.1
100.4
98.4
100.3
95.0
71.1
77.6
96.8
100.9
100.3
102.2
100.0
100.2
100.0
100.1
134.4
Table 2A. Producer, consumer and shared responsibility by columns, Gg of CO2
Year 2000
(A)- 01 Agriculture, stockbreeding and
hunting
(A)- 02 Silviculture and forestry exploitation
(B)- Fishing
(C)- 10 Anthracite, coal, lignite and peat
(C)- 11 Oil crudes, natural gas and uranium
(C)- 13 Metallic minerals extraction
(C)- 14 Non-Metallic minerals extraction
(D)- 15 Food and beverages
(D)- 17 Textile industry
(D)- 18 Clothing and furs
(D)- 19 Leather and footwear
(D)- 20 Wood & products of wood & cork
(D)- 21 Pulp, paper & paper products
(D)- 22 Printing & publishing
(D)- 23 Mineral oil refining, coke & nuclear
fuel
(D)- 24 Chemicals
(D)- 25 Rubber & plastics
(D)- 26 Non-metallic mineral products
(D)- 27 Basic metals
(D)- 28 Fabricated metal products
(D)- 29 Mechanical engineering
(D)- 30 Office machinery
(D)- 31 Electric machinery and materials
(D)- 32 Electronic materials
(D)- 33 Medical, precision & optical
instruments
(D)- Motor vehicles, trailers & semi-trailers
(D)- 35 Other transport equipment
(D)- 36 Furniture, miscellaneous
manufacturing
(D)- 37 Recycling
(E)- 40 Electricity, gas and water supply
(E)- 41 Water collecting, treatment &
distribution
(F)- 45 Construction
(G)- 50-52 Motor vehicles and reparation
(H)- 55 Hotels & catering
(I)- 60 Inland transport
(I)- 61 Water transport
(I)- 62 Air transport
(I)- 63 Supporting & auxiliary transport
activities
(I)- 64 Communications
(J)- 65-67 Financial intermediation
(K)- 70-74 Real estate activities & business
services
(L)- 75 Public administration
(M)- 80 Education
Year 2005
SRcols/
PRcols CRcols
PRrows (%)
SRcols
SRcols/
PRrows (%)
4,031
55
3,658
85
91
4
70
25,848
1,730
3,481
1,675
181
958
1,095
4,964
60
2,892
84
82
70
180
23,911
1,711
2,626
1,493
259
1,414
1,028
60
59
92
8
23
40
36
477
120
1974
1234
58
53
519
12,266
9,976
1,177
11,158
6,909
2,606
3,132
237
1,821
577
22,641
12,535
379
2,162
-860
2,882
8,799
1,469
1,936
4,247
21,055
12,547
743
5,873
1,693
2,928
7,113
1,070
2,002
3,065
104
182
581
13
13
739
2280
15280
2043
34051
12477
4346
1854
406
7,394
1,512
1,817
17,772
3,091
1,370
14,778
2,672
19568
6657
3930
1,639
125
80,648
735
39
89
1,676
-3
23,.929
3,843
-5
44,061
3,244
-4
40,558
1455
-1
45
863
51,354
21,766
18,600
8,303
632
3,647
623
7,845
9,145
5,396
16,354
1,748
5,779
114
296
175
194
84
64
78
757
52,148
21,577
15,394
10,679
2,715
6,873
1,176
70,966
31,608
22,744
8,977
1,010
5,827
1,095
66,660
30,382
21,307
10,344
1,608
5,995
201
2512
580
766
53
59
81
1,755
255
228
2,284
2,222
1,915
1,814
668
654
117
288
314
3,230
2,178
1,290
3,463
3,573
1,965
3,643
3,329
1,871
235
1435
899
624
817
596
12,125
9,817
4,511
3,226
2,541
1,374
569
342
253
10,351
8,091
3,076
15,162
14,303
5,293
14,488
13,113
4,874
2555
1765
899
PRcols CRcols
SRcols
8,660
112
2,563
842
295
204
565
6,158
1,750
157
141
529
3,316
224
3,263
61
3,506
8
443
-6
25
19,231
1,225
3,574
1,177
115
188
1,214
7,727
86
3,644
820
491
77
451
10,240
1,803
1,649
702
484
2,096
673
93
85
116
76
136
43
91
204
127
1240
580
109
78
340
5,238
53
1,834
70
28
160
337
17,437
1,445
831
922
333
1,972
731
20,572
8,294
153
50,260
14,205
465
566
8
116
10
17,765
8,617
199
1,245
135
2,365
6,614
2,080
1,800
2,918
20,077
9,266
639
38,461
7,317
1,428
4,451
1,266
1,426
2,018
99
134
499
85
55
361
1427
18081
1455
22419
8
249
84
1,525
13,092
1,807
873
9,649
1,260
242
370
92,445
3,762
0
38,163
679
3,008
5,878
3,134
21,921
3,580
7,455
32
(N)- 85 Health and social work
(O)- 90-93 Other community, social &
personal serv.
(P)- 95 Priv. households with employed
persons
Total
1,010
6,885
2,162
236
5,677
9,572
8,828
965
1,134
7,197
2,235
223
5,436
8,742
8,114
808
0
248,692
0
0
-
0
0
0
288,231
273,050
110
265,637 374,113
-
357,133
144
Sources for Tables 1A and 2A: Own work from input-output tables and CSEA.
33