Directory UMM :Data Elmu:jurnal:E:Environmental Management and Health:Vol10.Issue3.1999:
Regional productive structure and water pollution in
the Ebro Valley (Spain)
Rosa Duarte Pac
Department of Economic Analysis, Faculty of Economics and Business Studies,
University of Zaragoza, Spain
Julio SaÂnchez-ChoÂliz
Department of Economic Analysis, Faculty of Economics and Business Studies,
University of Zaragoza, Spain
Keywords
Input/output analysis, Structure,
Pollution, Economics
Abstract
This paper uses the input-output
methodology in order to study the
water pollution associated with an
economic system, through an examination of the productive relationships which link the different
sectors and their involvement in
the water pollution that is generated. The indicators associated
with the demand models are obtained on the basis of the construction of primary pollution and
return matrices. In a second
stage, an analysis is made of the
trade-offs between economic and
environmental variables, as a way
of linking the variables to be taken
into account for the purposes of
environmental planning. The empirical application is made for the
regions of the Ebro Valley (Spain).
The results show the importance
of considering the responsibility of
the agriculture, livestock and food
sectors for the coherent treatment
of the water pollution problem.
Environmental Management
and Health
10/3 [1999] 143±154
# MCB University Press
[ISSN 0956-6163]
Introduction
Recognition has traditionally been given to
the impact of industrial or agricultural
activity over the environment and, more
specifically, with respect to the pollution of
continental waters. However, despite this
explicit recognition, the economic and environmental literature offer a very limited
number of references where attention is
given to the productive relationships that
link the different sectors of the economy.
Furthermore, within the framework of responsibilities for environmental pollution,
there are few studies that examine precisely
which economic sectors have a highly polluting production and, above all, which are
effectively providing incentives to pollute in
any given area by virtue of being the
purchasers of the inputs produced by other
productive sectors which have a significant
polluting effect. Logically enough, the challenge of improving the environment through
a control over production and the pollution
associated with each productive sector is
necessarily linked to a deeper knowledge of
the structural relationships between economic activities and the environmental conditions in which they are carried out.
Although we often make general reference
to the scarcity of water in specific areas,
closer examination shows that this problem
is intimately related to that of quality. In
other words, whilst there may well be
sufficient water in these areas, the pollution
generated by specific industrial or agricultural activities limits its uses. This form of
scarcity is due both to the polluting effects of
intensive agriculture, involving the massive
use of harmful fertilisers, as well as to other
factors, from amongst which we can place
emphasis on the increase and concentration
of the population in urban centres, growing
industrialisation and centralisation in industrial areas and the ever-increasing complexity of productive processes.
Against this background, a circulatory
model shows us how each agent or user is, in
turn, the supplier of water for subsequent
uses. Thus, each productive sector, or each
agent, is responsible for the alterations that
its activity causes (directly or indirectly) to
the quality of the resource. It is therefore
necessary to define the levels of responsibility of each user so that we can draw closer to
the nucleus of the problem of water pollution.
In this paper, and with reference to the
Ebro Valley, in the North-East of Spain[1],
our objective is to characterise the sectors of
an economy with respect to water pollution
from a dual perspective: first, by considering
the proposal of Rasmussen for the determination of ``key sectors'', but introducing some
qualifications with respect to the indicators
obtained; and with this analysis being completed with a brief description of the responsibilities of the different components of the
economic demand of the system; secondly, by
considering the relationships between economic and environmental variables through
what has come to be known as the study of
the trade-offs. Emphasis is placed on the
framework of economic relationships within
the region under study from a macro-economic point of view, one which considers
both intersectorial relationships and the
dependencies between groups and productive
processes.
The rest of the paper is organised as
follows. Section 2 is devoted to the antecedents, the initial assumptions, the criteria for
the aggregation of the sectors and the homogenisation of the tables and the characterisation of water pollution. In Section 3 we
adopt a theoretical approach towards obtaining the indicators derived from the inputoutput model. In a first analysis we refer to
the characterisation of key sectors following
Rasmussen, and subsequently to the obtaining of the trade-offs. In Section 4 we comment
on the results obtained from an application of
this methodology to the Ebro Valley. Section
[ 143 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
5 closes the paper with a review of the most
relevant conclusions.
Environmental Management
and Health
10/3 [1999] 143±154
Antecedents, sources and initial
assumptions
Antecedents
Ever since the pioneering work of Leontief in
1946, models based on input-output tables
have been used in all countries as instruments for estimation and economic forecasting. The explanatory power of these models
has also been taken advantage of in environmental modelling, given that they offer an
integrated view of the entire economy, with
attention being paid to all the relationships
that link the sectors. In this environmental
field, the first attempt to extend the traditional input-output models to encompass the
study of environmental pollution was made
on the part of Leontief himself, when in 1970
he proposed an integrated method to consider
productive activities and their environmental effects in the functioning of the economy.
The basis of environmental input-output
analysis is found in the input-output tables.
These tables describe the relationships between the sectors of an economy, taking into
account the resource flows between sectors,
the final products and the inputs or resources
used to produce them. The variables are
generally expressed in monetary terms. Its
extension to the environmental field concentrates on adding various additional tables to
the traditional economic tables in order to
describe the resource flows obtained from the
environment and the different types of pollution that appear as a consequence of the
productive activities. Following this line,
Daly (1980) offered a first approximation
which considered the functions carried out
by the environment (supplier of inputs,
supplier of environmental services and receiver of residues). This author's work supposed a starting point for the subsequent
construction of environmental accounts
linked to multisectorial models. In the earlier
mentioned interesting work on atmospheric
pollution in the USA, Leontief (1970) added
three tables to the traditional economic
input-output tables in an attempt to reflect
the relationships between the environmental
and economic sectors and between the environmental sectors themselves. Nevertheless, he himself recognised the impossibility
of obtaining data on these latter relationships, and in a first analysis, the flows are
recognised in monetary terms. Another important work is that of Victor (1972), who
proposed the construction of satellite accounts which, when united with the eco-
[ 144 ]
nomic accounts, would allow for a
description of the environment. This author
placed emphasis on the need to consider
environmental flows in physical terms. On
the basis of these developments, and principally on the work of Victor, the papers of
Stone (1980) and of Miller and Blair (1985)
offered ``hybrid'' models (with the economic
part expressed in monetary units and the
environmental part in physical units) for the
study of pollution.
Works that have employed the input-output technique for environmental analysis
have usually chosen to adopt what Pajuelo
has called a ``partial equilibrium approach'',
that is to say, to consider only those flows
that move from the economic to the environmental medium. In this regard, see the
interesting works of Pajuelo (1980), a pioneer
in Spain in the analysis of atmospheric
pollution using this methodology, and of
AlcaÂntara (1995) who contributes very novel
methods for the study of contamination from
CO2, NOX and SOX, some of which are
considered in this paper.
Sources and aggregation criteria
A study using the input-output methodology
begins by considering the interrelationships
between the productive sectors and the
environment. In our particular case, we are
interested in studying these relationships in
the Ebro Valley and in concentrating on
water pollution. We do this by adding to the
tables a number of additional rows that
reflect the direct unitary returns and the
direct unitary pollution for the main polluting agents, BDO5 (biological demand oxygen
in five days), TSS (total suspended solids),
nitrates and phosphates. This procedure has
been followed for the four economies of the
Valley, that is to say, Lerida-Tarragona,
AragoÂn, Navarra and La Rioja, as well as for
the aggregation of all these areas (the Ebro
Valley). Although we are conscious of the fact
that water pollution cannot be totally characterised by these four indicators, they do
nevertheless offer a first synthesis of the
agricultural and industrial pollution (supposing that the regulations on toxic and
dangerous residues are respected). The first
two indicators are usually employed as a
guide to determine the physical size of water
purification plants in urban centres and for
drawing-up the scale of charges in order to
recover part of the purification costs. In turn,
it is undoubtedly the case that the nitrates
and phosphates used in the agriculture and
livestock sectors are mainly responsible for
rendering water useless for subsequent purposes.
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
The multisectorial descriptions have been
obtained from the regional tables corresponding to each area. The study has been
carried out for the following 21 representative sectors: cereals under irrigation, industrial crops, vegetables, fruit, other crops
under irrigation, dry land crops, livestock,
energy, metals, non-metals, chemicals, motor
vehicles, dairy products and juices, wine
production, the rest of the food industry,
manufacturing, paper, construction, hotel
and catering, health and, finally, other services. These sectors have been selected by
paying attention to economic (demand and
production), functional (similarity in the
productive activity), and environmental
(sectors with an important weight in the
consumption, returns or direct pollution of
water) criteria. Some aggregation and sector
selection criteria, as well as the original
tables, can be found in SaÂnchez-ChoÂliz (1997)
and in SaÂnchez-ChoÂliz and Duarte (1997).
For the unitary returns and pollution
coefficients, we have used as a starting point
the estimations of sectorial water pollution
compiled by World Health Organisation in
the WHO offset publication No. 62 (1982).
These estimations have been adjusted to the
reality of the area under study by using other
regional sources such as the industrial data
base on water demand produced by the Ebro
Water Confederation (CHE), as well as employment and production statistics, etc. Similarly, for the coefficient of the agricultural
sector, account has been taken of the average
crop distributions in the area under study,
considered as at 1992, with this information
being taken from the Agrarian Statistics
Yearbook, produced by the Spanish Ministry
of Agriculture, Fisheries and Food (MAPA,
1993), as well as from studies on the water
needs of the different crops (Tabuenca, 1994)
and on irrigation efficiency (Bielsa, 1998), etc.
The base regional input-output tables have
been updated by using the RAS method. On
the basis of the tables for AragoÂn (1992),
Navarra (1995), La Rioja (1974) and CatalunÄa
(1987), all updated to 1992 and homogenised
into the earlier-mentioned 21 sectors, we
have drawn up the aggregate tables for the
entire Ebro Valley, which we have used for
the analysis carried out in the Sections 4 and
5.
Finally, we should recall that the use of
input-output models supposes that we first
assume the linearity of the production function and the constancy of the technical
coefficients (assumptions that have often
been the subject of criticism). Additionally,
we must assume the constancy of the returns
and pollution coefficients and the lack of
pollution on the part of the river itself. This
supposes that we take as a starting point the
self-purifying capacity of the river in the
absence of productive activity and, therefore,
we assume that each industry receives unpolluted water and that the pollution found
downstream of its activity corresponds only
to that industry.
The theoretical model
In an economy made up of n productive
sectors, where Xj represents the effective
production of sector j, Yj the destination of
the Xj to the final demand and Xi the use that
sector j makes of the production of sector i ,
we have for any sector j:
Xj Xj1 Xj2 . . . Xjn Yj
Accepting the hypothesis that the proportion
of factors used by each sector is fixed, the
technical coefficients aij can be defined as:
aij Xij=Xj
For all the sectors of the economy and in a
matrix form, the system can be expressed as:
X AX Y
1
where X is the productions vector, Y is the
final demand vector and A is the technical
coefficients matrix.
Let us define a matrix C, which a matrix of
direct unitary pollutants where each element
ckj represents the unitary returns (for k = 1)
and the direct unitary pollution of each type
produced in sector j. The direct unitary
returns are expressed in m3/million pesetas,
whilst the pollutants are expressed in kg/
million pesetas.
If we pre-multiply expression (1) by vector
C, we obtain:
E CX CAX CY
2
where E is the matrix of total returns and
pollutions of the economy.
Expression (1) can also be written as:
X I ÿ Aÿ1 Y
3
where I ÿ Aÿ1 is the Leontief inverse matrix. The total by columns of the ij elements
of the inverse provide us, for each sector,
with the direct and indirect input requirements when the final demand of each sector
increases by one unit.
If we pre-multiply by C, we obtain:
VC CX C I ÿ Aÿ1 Y
4
We call the generic element of this matrix,
VCkj , the ``valuation in k-type pollution'',
where this represents the k-type pollution
directly or indirectly generated by the sector
j in obtaining its final demand.
[ 145 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Furthermore, in unitary terms we can
propose the following indicators:
Environmental Management
and Health
10/3 [1999] 143±154
with
vc C I ÿ Aÿ1
vckj
n
X
ckj ij
5
i1
is the matrix of ``pollution values'' associated
with the backward linkages or unitary coefficients of backward linkages, and which
shows the pollution directly and indirectly
generated by the sectors in obtaining one
unit of final demand. It measures the backward linkages in the sense that they are
indicators of the capacity of the different
sectors to ``drag'' the others to produce
resources, and therefore to pollute, with
these resources being used as inputs to
obtain one unit of final demand. The interpretation of these coefficients has to be made
in terms of ``values'' (the Marx value of
labour, or water value (SaÂnchez-ChoÂliz et al.,
1994)). The interpretation of these co-efficients as values opens an avenue for us to
compare these with the monetary values of
specific types of economic activities. Similarly, these coefficients are interpreted by
Pasinetti (1977) as vertical integration coefficients.
We can also construct other indicators,
paying attention to the role of the different
sectors as suppliers of inputs for other
productive sectors. In this case, the relevant
matrix is not the technical coefficients matrix, but rather the B matrix of distribution
coefficients. These coefficients are called the
forward linkage coefficients and measure the
capacity of each sector to impel the production of other sectors, in the sense of the sales
that this sector makes to the other sectors in
the economy. Let us describe the unitary
forward linkage in the pollution as:
n
X
ij
6
icki
j1
which is the returns or pollution generated
by the ith sector (of type k) in the face of an
increase of one unit in the demand of all the
sectors. Pajuelo (1980) described it in terms of
the forward linkage potential in the emission
and it is a measure of the sensitivity of each
sector to be forced by the economy to
generate a specific type of pollutant (given
that the other sectors demand its products as
inputs).
In the particular case we are analysing
here, we can see from the pollution values
that there is similar behaviour in the different regions on the part of particular sectors.
[ 146 ]
On the basis of such indicators, Pajuelo
(1980) adapts those constructed by Rasmussen (1956) in order to determine the key
sectors and proposes the relative backward
and forward linkage indicators in pollution.
For the case of water pollution, and following
Pajuelo's terminology, we have defined the
following two coefficients:
j : the relative backward linkage coefficient in pollution. This shows the importance
that the returns and pollution of sector j have
in the total pollution that can be assigned to
the productive sector. It allows us to see the
polluting role played by one sector, in that it
incorporates the inputs of the other sectors
(that pollute) in its productive process.
j
1
n2
1
n vckj
n
P
7
vckj
j1
In an analogous manner, we can construct
the relative forward linkage coefficients (i)
that demonstrate the extent to which the total
returns of one sector vary in the face of an
increase in the final demand of all the sectors
in relation to the average behaviour of the
productive system.
i
1
n2
1
n icki
n
P
8
icki
i1
Following the idea of Rasmussen (1956),
Pajuelo (1980) establishes a classification,
using the forward and backward linkage
coefficients, adapted to atmospheric pollution (see Table I).
The key sectors are identified as those with
a marked polluting character as the demander of inputs and as the suppliers of resources to other sectors. In this sense, any
pollution-reducing activity in these key sectors that might affect their production could
have serious consequences for the whole
productive system.
However, it must be said that the use of
these indicators to determine key sectors has
been heavily criticised, in that they do not
provide information on their degree of
Table I
Classification using the forward and backward
linkages of coefficients
mj > 1
mj < 1
mj > 1
Key sectors
mj < 1
Sectors that
exhibit backward
linkage in the
pollution
Sector that exhibits
forward linkage
in the pollution
Other sectors
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
dispersion; that is to say, on the extent to
which the pollution generated by way of
intersectorial relationships is concentrated
in a reduced set of sectors or whether it
depends on the operation of almost all the
sectors of the economy. In order to examine
more deeply the information offered by the
relative backward and forward linkage coefficients, we can construct the backward and
forward linkage variation coefficients in
returns and pollution. These indicators try to
detect the level of pollution dispersion that
exists in the relationships of the economy. On
the basis of the indicators proposed by
AlcaÂntara (1995), we can, for any pollutant k
and for any sector j, construct the following
indicators:
Backward linkage variation coefficient in
type k pollution:
v
u
!2
uP
n
P
un
cki ij ÿ n1
cki ij
u
i1
1 u
ui1
CVAckj
u
n
P
n ÿ1t
1
cki ij
n
i1
Forward linkage variation coefficient in type
k pollution:
v
u
!2
u
n
n
P
P
u
1
c
c
ÿ
u ki
ij
ij
n ki
j1
j1
1 u
u
CVIcki
u
n
P
n ÿ 1t
1
ij
n cki
the economic sectors. The trade-offs approach and, on this basis, the extension of the
concept of elasticity, allows us to examine
these types of relationships more deeply.
The way to construct the trade-offs indicators between environmental and economic
variables can be found in AlcaÂntara (1995).
Having obtaining the valuation in k-type
pollution, we can similarly construct the
valuations in terms of income and employment
^ where v is the vector of
VV v I ÿ Aÿ1 y
direct unitary coefficients of added value,
^ where l is the vector of
and VL l I ÿ Aÿ1 y
direct unitary coefficients of labour. We can
also obtain the income-pollution and employment-pollution trade-offs as follows:
TOv ; ck VCk ÿ1 VV
TOL ; ck VCk ÿ 1 VL
These indicators are ranked from the largest
to the smallest. Sectors with both indicators
lower than the average are sectors where
reductions in pollution have a smaller effect
on the economic variables of income and
employment than in other sectors, which
means that the ``cost'', in terms of these
economic variables of choosing these sectors
for the reduction of pollution will have a
lower impact than in the other sectors of the
economy.
i1
Given the actual construction of the indicators, the backward linkage variation coefficients take high values when the return or
the pollution incorporated by a sector comes
from purchasing relationships with a reduced number of sectors. Here we are talking
of sectors whose polluting character is determined by their direct or indirect dependence on the purchases made from a small
number of highly polluting sectors within the
economy.
Similarly forward linkage variation coefficients take high values when the ``sales'' of
pollution made directly or indirectly by a
sector are concentrated in a reduced group of
sectors.
Trade-offs between pollution and economic
variables
In the search for better pollution indicators
that might assist us in determining which
are the key sectors of the economy with
respect to any type of pollution, Karunaratne
(1989) has proposed the trade-offs between
energy and economic variables. The idea is
that any measure for the reduction of pollution must take into account the repercussions in terms of income or employment in
Application to the Ebro Valley
Having considered the theoretical basis, let
us now turn to the results that we have
obtained for the Ebro Valley. Although our
analysis has been done to the four economic
structures of Lerida-Tarragona, AragoÂn, Navarra and La Rioja, given the amount and
complexity of the results, we only present
now the results for the aggregate table of the
Valley.
We start with the determination of the key
sectors proposed by Rasmussen (1956). Having obtained the backward and forward
linkages for all the sectors[2], we calculate
the relative coefficients, which will allow us
to draw the role played by each sector in the
economy. Table II and Table III list these
relative coefficients for the Ebro Valley.
On the basis of these results we can obtain
a classification, for each type of pollutant, as
shown in Table IV.
Thus, from the aggregated results for the
Valley as a whole, we can note how all the
irrigated agricultural sectors have a greater
relative backward linkage capacity with
respect to the volume of returns, followed by
the food sectors, livestock and paper. With
[ 147 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
respect to BDO5 pollution, the relative backward linkage sectors are metals, dairy products and wine production and livestock.
With respect to TSS pollution, it is the
energy, non-metal, paper and food sectors
that display the greatest backward linkage,
with the other sectors displaying values that
are similar amongst themselves but significantly lower than these first sectors. Finally,
with respect to nitrates and phosphates the
most relevant sectors are livestock (the great
value in both pollutants), followed by a part
of the food sector (mainly meat production)
and dairy products. Note that both sectors
are very joined to livestock production. After
these sectors the agricultural sectors are the
ones with a great relative backward linkage.
Similarly we can comment on the main
results obtained in the analysis of the relative forward linkages. With respect to the
volume of returns, the irrigated agricultural
sectors have the greatest values, followed by
chemicals and paper. In BDO5 the main
relative forward linkage is for dairy products
and juices, metals and livestock. In TSS the
classification is different: energy, non-metals, paper and chemicals are the sectors with
the greatest coefficient, although the food
industry also obtains great values. In nitrates
and phosphates only the livestock sector is
Table II
Relative backward linkages ± the Ebro Valley
Sectors
Cereal crops
under irrigation
Industrial crops
Vegetables
Fruits
Other crops
under irrigation
Dry land crops
and others
Livestock
Energy
Metals
Non-metals
Chemicals
Motor vehicles
Dairy products
and juices
Wine production
Other foods
Manufacturing
Paper
Construction
Hotels and
catering
Health services
Other services
Average
[ 148 ]
Returns
BDO5
TSS
Nitrates
Phosphates
6.87
0.11
0.14
0.49
0.46
2.59
2.01
1.86
4.15
0.11
0.12
0.11
0.12
0.14
0.14
0.14
0.14
0.54
0.52
0.53
0.50
0.54
0.51
0.52
0.48
0.03
0.12
0.14
0.55
0.55
0.49
0.02
0.04
0.14
0.39
0.02
0.65
0.85
0.03
1.47
0.09
0.12
0.15
15.56
0.13
2.10
0.35
7.23
1.72
0.31
3.09
13.67
0.00
0.00
0.00
0.01
0.00
1.46
13.73
0.00
0.00
0.00
0.01
0.00
1.47
0.48
0.45
0.07
0.46
0.03
0.17
1.16
0.20
0.12
0.20
0.10
0.19
1.60
0.26
0.65
1.42
0.81
0.31
0.10
2.25
0.05
0.11
0.00
0.19
0.10
2.26
0.05
0.11
0.00
0.19
0.04
0.04
1.00
0.03
0.02
1.00
0.07
0.10
1.00
0.01
0.01
1.00
0.01
0.01
1.00
representative in terms of relative forward
linkage.
If we pay attention to the classification
derived from the relative forward and backward linkages, the results for the Valley can
be compiled as follows: With respect to the
volume of returns, the key sectors are cereal
crops under irrigation, industrial crops,
fruits and other crops under irrigation, if we
consider all the sectors of the economy. If we
only consider the industrial sectors, the
chemical industry and paper sectors are the
most relevant in terms of relative forward
and backward linkage, although the food
sectors also have an important backward
linkage (due to the strong relationship with
agricultural production). Note that in LeridaTarragona and La Rioja, the chemical sector
does not appear to be important from the
forward linkage point of view.
As regards BDO5 pollution, the metal
sector, and food production sectors appear to
be key and only the livestock sector exhibits
forward linkage. This sector pushes the BDO5
pollution in the food sectors mainly.
For TSS pollution, the key sectors are
energy and non-metals, chemicals, agrifood
production and paper, although these results
are not common for all the regions.
Finally, with respect to the other pollutants, the livestock sector obtains the higher
values of these indicators in all the regions.
As a result, we can see that in all cases
there are three significant sectors, either as
key sectors or as sectors that exert backward
linkage over the others, thereby forcing
water pollution, namely the agriculture, livestock and agro-food sectors. Also note the
difference between sectors with respect to the
different types of pollutants. The importance
of the backward linkage in terms of BDO5 and
TSS (except for the livestock sector) is
concentrated fundamentally in the industrial
sectors (chemicals, paper, metal and motor
vehicles), whilst the responsibility for returns lies almost entirely with agriculture
and the sectors which depend on agricultural
production (agro-food, paper). The paper
sector stands out with respect to the volume
of returns. This is quite logical, given that
this sector receives inputs from the agricultural sector (a large-scale consumer and also
with important volumes of returns), whilst it
also supplies resources to almost all the
sectors of the economy. The chemical industry also appears as an important sector from
the point of view of water pollution,
especially in AragoÂn and Navarra. Both in
the different regions and in the Valley as a
whole, there appear to be no clear forward
linkages. In nitrates and phosphates the
livestock sector is the only one relevant.
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
Again we can note that the key sectors are
identified as those with a marked polluting
character as the demanders of inputs and as
the suppliers of resources to other sectors. In
this sense, any pollution-reducing activity in
these key sectors that might affect their
production could have serious consequences
for the whole productive system.
The next question we must raise concerns
the character of the ``sales and purchases'' of
pollution made by the sectors, that is to say,
the pollution incorporated by those sectors
with backward linkage potential, or introduced by those with forward linkage potential in their direct or indirect sales: does this
result from transactions with the majority of
the sectors in the economy or with just a
reduced number of these sectors? Table V
shows the coefficients of variation in the
backward and forward linkages:
The Table contains the results obtained for
the Ebro Valley. Note that it is the sectors
which reflect agriculture under irrigation
and basic industry, especially non-metals,
chemicals and paper, that obtain the highest
values for the variation coefficient in backward linkage in terms of returns. This means
that the returns which these sectors incorporate into their production have their origin
Table III
Relative forward linkages ± the Ebro Valley
Sectors
Cereal crops
under irrigation
Industrial crops
Vegetables
Fruits
Other crops
under irrigation
Dry land crops
and others
Livestock
Energy
Metals
Non-metals
Chemicals
Motor vehicles
Dairy products
and juices
Wine production
Other foods
Manufacturing
Paper
Construction
Hotels and
catering
Health services
Other services
Average
Returns
BDO5
TSS
Nitrates
Phosphates
7.57
0.00
0.00
0.04
0.00
2.73
2.20
1.97
4.19
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.39
0.02
0.06
0.14
0.45
0.01
0.19
1.23
0.00
2.40
0.03
0.12
0.07
15.62
0.00
3.18
0.44
8.32
1.81
0.20
2.99
20.91
0.00
0.00
0.00
0.00
0.00
0.00
21.00
0.00
0.00
0.00
0.00
0.00
0.00
0.24
0.06
0.08
0.48
0.01
0.10
1.16
0.03
0.14
0.19
0.00
0.01
1.43
0.22
0.81
1.45
0.00
0.14
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.03
0.07
1.00
0.00
0.00
1.00
0.00
0.00
1.00
0.00
0.00
1.00
0.00
0.00
1.00
in the inputs which are directly or indirectly
incorporated from a reduced number of
sectors. In general, the agricultural sectors
incorporate returns in their own self-consumption, whilst the industrial sectors produce their returns partly through their own
direct contribution and partly from the
inputs coming from other sectors (in the case
of paper, probably from the agricultural
sectors). With respect to the variation in the
forward linkages, we can note that it is the
cereal crops under irrigation sector which
obtains the highest value of the coefficient,
showing that the returns directly or indirectly incorporated by this sector (which, in
relation to the average for the economy, is
high), are concentrated in a very limited
number of sectors (mainly self-consumption
and the agrifood industries). Both with
respect to forward and backward linkages, it
is the final industry and services sectors that
maintain the most homogenous relations
with all the sectors, in such a way that the
returns they incorporate into their own
production or into the production of other
sectors is made via the transactions they
carry out with the majority of the sectors in
the economy.
Turning now to BDO5 pollution, it is
interesting to note that the highest values of
the variation coefficients in backward linkages are obtained in a number of groups:
first, in the livestock and energy sectors
(essentially by way of their direct pollution);
secondly, in the agrifood industry sectors
(fundamentally because of their purchasing
relationships with the livestock sector); and
finally, in the paper and construction sectors,
with both of these being large purchasers of
inputs from the energy sectors. Thus, we can
say that a significant incorporation of BDO5
results from the relationships that some
primary sectors (large direct polluters), specifically energy and livestock, maintain with
their most important purchasers of inputs.
The variation coefficients in backward linkages in BDO5 show that it is the energy
sector which directly or indirectly incorporates a large volume of pollution to a small
group of sectors. The observations made with
respect to BDO5 can be repeated in very
similar terms with respect to TSS pollution,
save for the livestock sector, which looses
importance in this regard. In the case of both
BDO5 and TSS, it is the services sectors
which show the greatest dispersion, followed
by the final industry sectors (motor vehicles,
manufacturing), confirming that the polluting character of these sectors is due to the
numerous contributions of pollution they
receive via the inputs which they incorporate
from almost all the sectors of the economy.
[ 149 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
Thus, any measure aimed at reducing BDO5
or TSS pollution must, in principle, be
directed towards controlling the pollution
that is emitted by the sectors with the highest
concentration of forward or backward linkages, given that in this way, and by acting
against just a few sectors, the majority of the
pollution in the system will be controlled.
As regards nitrates and phosphates, all the
sectors obtain similar values for the variation coefficients in backward linkages. As we
can see from the earlier tables, this type of
pollution is, in direct terms, concentrated in
the agriculture and livestock sectors, particularly in the second of these, and to some
extent in the hotels and catering sector.
Thus, the pollution which is directly and
indirectly generated by each sector of the
economy will only be due to the relationships
maintained with these sectors. In the variation in forward linkages, we can see that it is
the livestock sector which incorporates its
pollution in the sales that it makes to a larger
number of sectors, whilst agriculture concentrates its pollution to a greater extent
(which is logical, given that the livestock
sector sells inputs not only to the agrifood
industry ± dairy and meat products ±, but
also to the hotel and catering and to the
manufacturing sectors).
In summary, we can say that for the area
under study, and save for the differences
which exist for each type of pollutant, the
pollution that is directly and indirectly
generated in the system is not, in general
terms, spread homogeneously. Rather, it is
Table IV
Classification of pollutants
Pollutant
Key sectors
Returns
BDO5
Nitrates
Irrigated crops
Metals, dairy
products, wine,
Energy, nonmetals,
chemicals,
dairy
products,
wine,
paper
Livestock
Phosphates
Livestock
TSS
[ 150 ]
Sectors that
exhibit backward
linkage in the
pollution
Sectors that
exhibit forward
linkage in the
pollution
None
None
None
Livestock
None
None
Dairy products
and juices,
other foods
Dairy products
and juices,
other foods
None
None
concentrated in the sales made by a limited
number of highly polluting sectors (fundamentally, livestock and primary industries)
to other final goods sectors (agrifood, paper,
construction) which make the biggest purchases of inputs. The results describe an
economy that is not particularly integrated, a
panorama which, in terms of pollution, can
be translated into very specific groups of
large-scale polluters and pollution receivers.
Following this line, the last stage is to
analyse the trade-offs between economic and
environmental variables. We have already
said that any environmental planning measures which seek to control water pollution
must be directed towards these polluting
``groups'', given that a large part of the
pollution associated with the system would
thereby be controlled. At this point, we
should recall that any pollution control
measure which has an impact on the production of the sectors must take economic
aspects into account, specifically the total
contribution made by these sectors to the
income and employment of the area in
question. It is necessary to place the returns
or the total pollution incorporated into a
sector in relation with the total income and
employment it generates. AlcaÂntara (1994)
interprets the trade-offs with an opportunity
cost measurement, in terms of income or
employment, which introduces anti-pollution
measures that might have an impact on the
demand and production of the sectors. The
results derived from this trade-off technique
are set out in Table VI.
The lowest values of the trade-offs are
obtained in almost all cases for the agriculture, livestock and agrifood industry, save in
the case of TSS, a form of pollution where the
lowest trade-offs correspond to non-metals,
followed by dairy products, chemicals and
paper. In all cases the highest values are
obtained in the services sectors, generally
followed by motor vehicles, energy and construction. The results obtained for agriculture and livestock are easily understood:
these are sectors in which the income or
employment that is generated per unit of
pollution directly or indirectly incorporated
into production is lower than in the other
sectors or, put another way, the ``profitability'' obtained from the incorporation of
pollution is lower. At this point, it is interesting to focus our attention on the results
obtained for the food industry. Here, the low
trade-off obtained in almost all the cases
shows us that we are considering a sector in
which the pollution that is incorporated
directly and, especially, indirectly (forcing
the production of polluting outputs in the
agricultural or livestock sectors), does not
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
``compensate'' for the generation of income or
employment in this sector. In these sectors,
any activities designed to reduce pollution
have a low ``opportunity cost'', that is to say,
they are sectors in which a reduction in
pollution (which could be associated with a
reduction in the demand or production of
these sectors), will have lower effects over
the income and employment of the economy
than any activities taken in other sectors. In
the case of BDO5 or TSS pollution, we can see
how the livestock and food sectors are united
with the metals, non-metals, chemicals and
paper sectors. As regards pollution caused by
nitrates and phosphates, the livestock sector
and the related food sector (dairy products
and meats) obtain the lowest values.
If we now relate these results with those
obtained earlier, we get a clearer idea of the
distribution of returns and pollution in our
economy and can say that, when studying the
water pollution in the Ebro Valley, we
encounter a number of well defined returns
and pollution ``distribution groups''. In the
case of returns, the key sectors are those
which comprise agriculture under irrigation.
These sectors have a very high volume of
direct returns and their self-consumption of
inputs is important. The returns generated
by the agriculture sector are either returned
directly or incorporated into the production
of a reduced number of sectors (some into
livestock, some into hotel and catering, some
into the paper sector). In the paper sector,
although the direct volume of returns is
important, its character is marked by its
dependence on the raw materials from the
agriculture sector. The income and employment that this sector generates is important,
so it would be reasonable in this case to
control the returns via a control over agricultural activities (for example, by introducing measures for the reuse of irrigation
water, the improvement of irrigation systems, etc.). In the case of BDO5 we can
identify a highly polluting sector, namely
livestock, which introduces pollution into the
system through the sales that it makes to the
food and hotel and catering sectors, etc.
Together with this sector, primary industry
(metals, non-metals) also shows important
forward and backward linkages. Therefore,
in this case we are faced with two large
blocks of polluting sectors: those related most
directly with livestock and those related with
basic industry (metals, non-metals, chemicals). The study of the trade-offs shows us
that the profitability obtained from pollution
Table V
Variation coefficients in the Ebro Valley
Sector*
Returns
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Average
4.5654
4.5207
4.5081
4.4979
4.5596
1.6948
2.5524
2.8211
3.5375
4.1507
4.4171
2.0851
1.8840
2.3388
2.1428
3.5411
4.3346
2.1102
2.4812
3.1447
3.3124
3.2953
Variation coefficients in backward linkages
TSS
Nitrates
Phosphates
BD05
3.4173
3.2675
3.3909
3.2637
3.4431
3.3181
4.3631
4.1795
4.5650
3.1917
3.6919
3.0147
4.5424
4.4010
3.0328
3.6182
4.0326
4.2117
2.9039
2.7464
2.9971
3.5996
2.6387
2.4042
2.3748
2.4690
2.5064
2.3049
2.0768
4.4650
3.5235
4.4963
4.0678
2.9231
4.3820
3.9187
2.7454
3.8996
4.2545
4.1470
2.0441
2.2889
2.1895
3.1486
4.2636
4.5069
4.5214
4.5295
4.3711
4.5183
4.5822
4.5591
4.5701
4.5661
4.5715
4.5735
4.5744
4.5201
4.5787
4.5808
4.5792
4.5664
4.5441
4.5758
4.5688
4.5344
1.0239
1.0245
1.0246
1.0246
1.0242
1.0245
1.0247
1.0234
1.0240
1.0238
1.0243
1.0242
1.0247
1.0245
1.0247
1.0246
1.0247
1.0238
1.0224
1.0246
1.0241
1.0242
Returns
4.1283
0.0389
0.0425
0.0456
0.0264
0.0000
0.0855
0.4214
0.2160
0.1614
0.0908
0.7320
0.1504
0.1337
0.2246
0.1868
0.0919
0.5069
0.2027
0.4187
0.1418
0.4023
Variation coefficients in forward linkages
BD05
TSS
Nitrates
Phosphates
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0249
8.9242
0.0171
0.1788
0.0913
0.1333
0.0086
0.0311
0.1651
0.0751
0.0745
0.0000
0.3728
0.0000
0.0000
0.8414
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0118
0.0308
0.0085
0.0181
0.0596
0.0153
0.0218
0.0466
0.0241
0.0210
0.0000
0.0681
0.0000
0.0000
0.0296
4.1283
0.2707
0.2981
0.3264
0.1766
0.2581
0.0047
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.6959
0.0000
0.0000
0.7699
0.9231
0.7681
0.8458
0.9264
0.5013
0.7325
0.0014
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.3504
0.0000
0.0000
0.6311
Notes:
* 1 = cereal crops under irrigation; 2 = industrial crops; 3 = vegetables; 4 = fruits; 5 = other crops under irrigation; 6 = dry land crops and others;
7 = livestock; 8 = energy; 9 = metals; 10 = non-metals; 11 = chemicals; 12 = motor vehicles; 13 = dairy products and juices; 14 = wine
production; 15 = other foods; 16 = manufacturing; 17 = paper; 18 = construction; 19 = hotels and catering; 20 = health services; 21 = other
services
[ 151 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
in industry is higher than that obtained in
the livestock-food complex. Thus, this group
combines its high polluting potential with
the lower costs, in terms of income and
employment, of acting against it by way of
measures for the reduction of pollution. This
picture changes somewhat when we consider
TSS pollution. In this type of typically
industrial pollution, it is the energy, nonmetals and chemicals sectors that now play
the key role, although the food industry also
takes relevant values. Finally, as regards
nitrates and phosphates pollution, it is the
livestock sector which appears as key according to the Rasmussen classification,
whilst the food industry sectors show a
marked backward linkage potential in pollution. Once again, it is the livestock-food
complex which concentrates the highest
productive incorporation of pollution, with it
also showing the lowest trade-offs.
Conclusions
In this work, we have applied some of the
tools made available by the input-output
methodology to the study of water pollution
in the Ebro Valley. We have offered no more
than a limited analysis, one which only
purports to demonstrate some of the possibilities that can be found in the study of the
relationships between the economy and the
environment. Notwithstanding this qualification, our initial analysis allows us to draw
the following conclusions.
The proposed analysis demonstrates the
possibility of using the traditional economic
models for a more in-depth study of water
pollution associated with the economic
structure of a region. In this paper we have
used two methods to approximate the relationships between economic and environmental variables. The first is an application
of the forward and backward linkage coefficients proposed by Rasmussen (1956) and
adapted by Pajuelo (1980), Proops et al. (1993)
and AlcaÂntara (1995) to the case of atmospheric pollution, whilst the second is an
approximation of the so-called trade-offs
relationships between economic and water
pollution variables. This paper represents a
first attempt to construct both types of
indicators for the study of the water pollution
Table VI
Trade-offs in the Ebro Valley
Sectors*
Returns
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Average
0.09
0.25
0.31
0.34
0.10
18.93
0.95
47.67
10.04
4.75
1.19
18.86
0.78
1.20
0.96
8.13
1.04
18.93
3.92
21.92
21.23
8.65
Trade-offs added value-pollutiona
BD05
TSS
Nitrates
1.39
1.50
1.36
1.52
1.33
1.42
0.15
7.20
0.08
1.91
1.00
0.64
0.01
0.13
0.58
1.32
0.63
1.70
0.91
7.57
10.95
2.06
13.38
14.50
14.16
14.30
13.65
14.65
11.24
1.17
4.06
0.28
0.87
3.61
0.54
1.17
5.41
2.96
1.07
2.57
6.66
37.58
26.50
9.06
0.20
0.19
0.20
0.20
0.20
0.19
0.01
74.39
25.00
36.19
10.73
16.22
0.06
0.94
0.03
1.95
0.74
29.68
0.57
8.84
22.68
10.91
Phosphates
Returns
0.35
0.32
0.33
0.33
0.34
0.32
0.01
121.94
40.92
59.26
17.57
26.54
0.09
1.56
0.05
3.18
1.21
48.61
0.93
14.45
37.13
17.88
0.03
0.08
0.11
0.12
0.04
5.96
0.51
5.21
2.50
1.00
0.30
3.69
0.26
0.28
0.35
2.98
0.25
6.46
1.31
6.38
7.25
2.15
Trade-offs employment-pollutionb
BD05
TSS
Nitrates
0.54
0.49
0.46
0.52
0.49
0.45
0.08
0.79
0.02
0.40
0.25
0.12
0.00
0.03
0.21
0.49
0.15
0.58
0.30
2.20
3.74
0.59
5.23
4.75
4.81
4.86
5.05
4.61
6.08
0.13
1.01
0.06
0.22
0.71
0.18
0.27
1.96
1.09
0.26
0.88
2.22
10.94
9.05
3.07
0.08
0.06
0.07
0.07
0.07
0.06
0.00
8.13
6.23
7.61
2.66
3.17
0.02
0.22
0.01
0.71
0.18
10.13
0.19
2.57
7.74
2.38
Phosphates
0.14
0.11
0.11
0.11
0.13
0.10
0.00
13.33
10.19
12.47
4.36
5.19
0.03
0.36
0.02
1.17
0.29
16.58
0.31
4.21
12.68
3.90
Notes:
a
Returns in thousands of pesetas/cubic metre, pollution in thousands of pesetas/kg of pollutant
b
Returns in thousands of jobs/cubic hectometre, pollution in thousands of jobs/kg of pollutant
* 1 = cereal crops under irrigation; 2 = industrial crops; 3 = vegetables; 4 = fruits; 5 = other crops under irrigation; 6 = dry land crops and others;
7 = livestock; 8 = energy; 9 = metals; 10 = non-metals; 11 = chemicals; 12 = motor vehicles; 13 = dairy products and juices; 14 = wine
production; 15 = other foods; 16 = manufacturing; 17 = paper; 18 = construction; 19 = hotels and catering; 20 = health services; 21 = other
services
[ 152 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
associated with a specific economic system,
in our case, the Ebro Valley in Spain, with
attention being given to a series of pollutants
that are characteristic of the agricultural,
livestock and industrial activities carried out
in that system. The results obtained from the
first analysis show that, in general terms, it
is the agriculture, livestock and agrifood
sectors which directly and indirectly emit
the highest volumes of returns and of nitrate
and phosphates pollution and for which the
``opportunity costs'' are lower in terms of
added value and employment. This situation
extends to the paper sector (also in terms of
volume of returns) and to the metals and nonmetals
the Ebro Valley (Spain)
Rosa Duarte Pac
Department of Economic Analysis, Faculty of Economics and Business Studies,
University of Zaragoza, Spain
Julio SaÂnchez-ChoÂliz
Department of Economic Analysis, Faculty of Economics and Business Studies,
University of Zaragoza, Spain
Keywords
Input/output analysis, Structure,
Pollution, Economics
Abstract
This paper uses the input-output
methodology in order to study the
water pollution associated with an
economic system, through an examination of the productive relationships which link the different
sectors and their involvement in
the water pollution that is generated. The indicators associated
with the demand models are obtained on the basis of the construction of primary pollution and
return matrices. In a second
stage, an analysis is made of the
trade-offs between economic and
environmental variables, as a way
of linking the variables to be taken
into account for the purposes of
environmental planning. The empirical application is made for the
regions of the Ebro Valley (Spain).
The results show the importance
of considering the responsibility of
the agriculture, livestock and food
sectors for the coherent treatment
of the water pollution problem.
Environmental Management
and Health
10/3 [1999] 143±154
# MCB University Press
[ISSN 0956-6163]
Introduction
Recognition has traditionally been given to
the impact of industrial or agricultural
activity over the environment and, more
specifically, with respect to the pollution of
continental waters. However, despite this
explicit recognition, the economic and environmental literature offer a very limited
number of references where attention is
given to the productive relationships that
link the different sectors of the economy.
Furthermore, within the framework of responsibilities for environmental pollution,
there are few studies that examine precisely
which economic sectors have a highly polluting production and, above all, which are
effectively providing incentives to pollute in
any given area by virtue of being the
purchasers of the inputs produced by other
productive sectors which have a significant
polluting effect. Logically enough, the challenge of improving the environment through
a control over production and the pollution
associated with each productive sector is
necessarily linked to a deeper knowledge of
the structural relationships between economic activities and the environmental conditions in which they are carried out.
Although we often make general reference
to the scarcity of water in specific areas,
closer examination shows that this problem
is intimately related to that of quality. In
other words, whilst there may well be
sufficient water in these areas, the pollution
generated by specific industrial or agricultural activities limits its uses. This form of
scarcity is due both to the polluting effects of
intensive agriculture, involving the massive
use of harmful fertilisers, as well as to other
factors, from amongst which we can place
emphasis on the increase and concentration
of the population in urban centres, growing
industrialisation and centralisation in industrial areas and the ever-increasing complexity of productive processes.
Against this background, a circulatory
model shows us how each agent or user is, in
turn, the supplier of water for subsequent
uses. Thus, each productive sector, or each
agent, is responsible for the alterations that
its activity causes (directly or indirectly) to
the quality of the resource. It is therefore
necessary to define the levels of responsibility of each user so that we can draw closer to
the nucleus of the problem of water pollution.
In this paper, and with reference to the
Ebro Valley, in the North-East of Spain[1],
our objective is to characterise the sectors of
an economy with respect to water pollution
from a dual perspective: first, by considering
the proposal of Rasmussen for the determination of ``key sectors'', but introducing some
qualifications with respect to the indicators
obtained; and with this analysis being completed with a brief description of the responsibilities of the different components of the
economic demand of the system; secondly, by
considering the relationships between economic and environmental variables through
what has come to be known as the study of
the trade-offs. Emphasis is placed on the
framework of economic relationships within
the region under study from a macro-economic point of view, one which considers
both intersectorial relationships and the
dependencies between groups and productive
processes.
The rest of the paper is organised as
follows. Section 2 is devoted to the antecedents, the initial assumptions, the criteria for
the aggregation of the sectors and the homogenisation of the tables and the characterisation of water pollution. In Section 3 we
adopt a theoretical approach towards obtaining the indicators derived from the inputoutput model. In a first analysis we refer to
the characterisation of key sectors following
Rasmussen, and subsequently to the obtaining of the trade-offs. In Section 4 we comment
on the results obtained from an application of
this methodology to the Ebro Valley. Section
[ 143 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
5 closes the paper with a review of the most
relevant conclusions.
Environmental Management
and Health
10/3 [1999] 143±154
Antecedents, sources and initial
assumptions
Antecedents
Ever since the pioneering work of Leontief in
1946, models based on input-output tables
have been used in all countries as instruments for estimation and economic forecasting. The explanatory power of these models
has also been taken advantage of in environmental modelling, given that they offer an
integrated view of the entire economy, with
attention being paid to all the relationships
that link the sectors. In this environmental
field, the first attempt to extend the traditional input-output models to encompass the
study of environmental pollution was made
on the part of Leontief himself, when in 1970
he proposed an integrated method to consider
productive activities and their environmental effects in the functioning of the economy.
The basis of environmental input-output
analysis is found in the input-output tables.
These tables describe the relationships between the sectors of an economy, taking into
account the resource flows between sectors,
the final products and the inputs or resources
used to produce them. The variables are
generally expressed in monetary terms. Its
extension to the environmental field concentrates on adding various additional tables to
the traditional economic tables in order to
describe the resource flows obtained from the
environment and the different types of pollution that appear as a consequence of the
productive activities. Following this line,
Daly (1980) offered a first approximation
which considered the functions carried out
by the environment (supplier of inputs,
supplier of environmental services and receiver of residues). This author's work supposed a starting point for the subsequent
construction of environmental accounts
linked to multisectorial models. In the earlier
mentioned interesting work on atmospheric
pollution in the USA, Leontief (1970) added
three tables to the traditional economic
input-output tables in an attempt to reflect
the relationships between the environmental
and economic sectors and between the environmental sectors themselves. Nevertheless, he himself recognised the impossibility
of obtaining data on these latter relationships, and in a first analysis, the flows are
recognised in monetary terms. Another important work is that of Victor (1972), who
proposed the construction of satellite accounts which, when united with the eco-
[ 144 ]
nomic accounts, would allow for a
description of the environment. This author
placed emphasis on the need to consider
environmental flows in physical terms. On
the basis of these developments, and principally on the work of Victor, the papers of
Stone (1980) and of Miller and Blair (1985)
offered ``hybrid'' models (with the economic
part expressed in monetary units and the
environmental part in physical units) for the
study of pollution.
Works that have employed the input-output technique for environmental analysis
have usually chosen to adopt what Pajuelo
has called a ``partial equilibrium approach'',
that is to say, to consider only those flows
that move from the economic to the environmental medium. In this regard, see the
interesting works of Pajuelo (1980), a pioneer
in Spain in the analysis of atmospheric
pollution using this methodology, and of
AlcaÂntara (1995) who contributes very novel
methods for the study of contamination from
CO2, NOX and SOX, some of which are
considered in this paper.
Sources and aggregation criteria
A study using the input-output methodology
begins by considering the interrelationships
between the productive sectors and the
environment. In our particular case, we are
interested in studying these relationships in
the Ebro Valley and in concentrating on
water pollution. We do this by adding to the
tables a number of additional rows that
reflect the direct unitary returns and the
direct unitary pollution for the main polluting agents, BDO5 (biological demand oxygen
in five days), TSS (total suspended solids),
nitrates and phosphates. This procedure has
been followed for the four economies of the
Valley, that is to say, Lerida-Tarragona,
AragoÂn, Navarra and La Rioja, as well as for
the aggregation of all these areas (the Ebro
Valley). Although we are conscious of the fact
that water pollution cannot be totally characterised by these four indicators, they do
nevertheless offer a first synthesis of the
agricultural and industrial pollution (supposing that the regulations on toxic and
dangerous residues are respected). The first
two indicators are usually employed as a
guide to determine the physical size of water
purification plants in urban centres and for
drawing-up the scale of charges in order to
recover part of the purification costs. In turn,
it is undoubtedly the case that the nitrates
and phosphates used in the agriculture and
livestock sectors are mainly responsible for
rendering water useless for subsequent purposes.
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
The multisectorial descriptions have been
obtained from the regional tables corresponding to each area. The study has been
carried out for the following 21 representative sectors: cereals under irrigation, industrial crops, vegetables, fruit, other crops
under irrigation, dry land crops, livestock,
energy, metals, non-metals, chemicals, motor
vehicles, dairy products and juices, wine
production, the rest of the food industry,
manufacturing, paper, construction, hotel
and catering, health and, finally, other services. These sectors have been selected by
paying attention to economic (demand and
production), functional (similarity in the
productive activity), and environmental
(sectors with an important weight in the
consumption, returns or direct pollution of
water) criteria. Some aggregation and sector
selection criteria, as well as the original
tables, can be found in SaÂnchez-ChoÂliz (1997)
and in SaÂnchez-ChoÂliz and Duarte (1997).
For the unitary returns and pollution
coefficients, we have used as a starting point
the estimations of sectorial water pollution
compiled by World Health Organisation in
the WHO offset publication No. 62 (1982).
These estimations have been adjusted to the
reality of the area under study by using other
regional sources such as the industrial data
base on water demand produced by the Ebro
Water Confederation (CHE), as well as employment and production statistics, etc. Similarly, for the coefficient of the agricultural
sector, account has been taken of the average
crop distributions in the area under study,
considered as at 1992, with this information
being taken from the Agrarian Statistics
Yearbook, produced by the Spanish Ministry
of Agriculture, Fisheries and Food (MAPA,
1993), as well as from studies on the water
needs of the different crops (Tabuenca, 1994)
and on irrigation efficiency (Bielsa, 1998), etc.
The base regional input-output tables have
been updated by using the RAS method. On
the basis of the tables for AragoÂn (1992),
Navarra (1995), La Rioja (1974) and CatalunÄa
(1987), all updated to 1992 and homogenised
into the earlier-mentioned 21 sectors, we
have drawn up the aggregate tables for the
entire Ebro Valley, which we have used for
the analysis carried out in the Sections 4 and
5.
Finally, we should recall that the use of
input-output models supposes that we first
assume the linearity of the production function and the constancy of the technical
coefficients (assumptions that have often
been the subject of criticism). Additionally,
we must assume the constancy of the returns
and pollution coefficients and the lack of
pollution on the part of the river itself. This
supposes that we take as a starting point the
self-purifying capacity of the river in the
absence of productive activity and, therefore,
we assume that each industry receives unpolluted water and that the pollution found
downstream of its activity corresponds only
to that industry.
The theoretical model
In an economy made up of n productive
sectors, where Xj represents the effective
production of sector j, Yj the destination of
the Xj to the final demand and Xi the use that
sector j makes of the production of sector i ,
we have for any sector j:
Xj Xj1 Xj2 . . . Xjn Yj
Accepting the hypothesis that the proportion
of factors used by each sector is fixed, the
technical coefficients aij can be defined as:
aij Xij=Xj
For all the sectors of the economy and in a
matrix form, the system can be expressed as:
X AX Y
1
where X is the productions vector, Y is the
final demand vector and A is the technical
coefficients matrix.
Let us define a matrix C, which a matrix of
direct unitary pollutants where each element
ckj represents the unitary returns (for k = 1)
and the direct unitary pollution of each type
produced in sector j. The direct unitary
returns are expressed in m3/million pesetas,
whilst the pollutants are expressed in kg/
million pesetas.
If we pre-multiply expression (1) by vector
C, we obtain:
E CX CAX CY
2
where E is the matrix of total returns and
pollutions of the economy.
Expression (1) can also be written as:
X I ÿ Aÿ1 Y
3
where I ÿ Aÿ1 is the Leontief inverse matrix. The total by columns of the ij elements
of the inverse provide us, for each sector,
with the direct and indirect input requirements when the final demand of each sector
increases by one unit.
If we pre-multiply by C, we obtain:
VC CX C I ÿ Aÿ1 Y
4
We call the generic element of this matrix,
VCkj , the ``valuation in k-type pollution'',
where this represents the k-type pollution
directly or indirectly generated by the sector
j in obtaining its final demand.
[ 145 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Furthermore, in unitary terms we can
propose the following indicators:
Environmental Management
and Health
10/3 [1999] 143±154
with
vc C I ÿ Aÿ1
vckj
n
X
ckj ij
5
i1
is the matrix of ``pollution values'' associated
with the backward linkages or unitary coefficients of backward linkages, and which
shows the pollution directly and indirectly
generated by the sectors in obtaining one
unit of final demand. It measures the backward linkages in the sense that they are
indicators of the capacity of the different
sectors to ``drag'' the others to produce
resources, and therefore to pollute, with
these resources being used as inputs to
obtain one unit of final demand. The interpretation of these coefficients has to be made
in terms of ``values'' (the Marx value of
labour, or water value (SaÂnchez-ChoÂliz et al.,
1994)). The interpretation of these co-efficients as values opens an avenue for us to
compare these with the monetary values of
specific types of economic activities. Similarly, these coefficients are interpreted by
Pasinetti (1977) as vertical integration coefficients.
We can also construct other indicators,
paying attention to the role of the different
sectors as suppliers of inputs for other
productive sectors. In this case, the relevant
matrix is not the technical coefficients matrix, but rather the B matrix of distribution
coefficients. These coefficients are called the
forward linkage coefficients and measure the
capacity of each sector to impel the production of other sectors, in the sense of the sales
that this sector makes to the other sectors in
the economy. Let us describe the unitary
forward linkage in the pollution as:
n
X
ij
6
icki
j1
which is the returns or pollution generated
by the ith sector (of type k) in the face of an
increase of one unit in the demand of all the
sectors. Pajuelo (1980) described it in terms of
the forward linkage potential in the emission
and it is a measure of the sensitivity of each
sector to be forced by the economy to
generate a specific type of pollutant (given
that the other sectors demand its products as
inputs).
In the particular case we are analysing
here, we can see from the pollution values
that there is similar behaviour in the different regions on the part of particular sectors.
[ 146 ]
On the basis of such indicators, Pajuelo
(1980) adapts those constructed by Rasmussen (1956) in order to determine the key
sectors and proposes the relative backward
and forward linkage indicators in pollution.
For the case of water pollution, and following
Pajuelo's terminology, we have defined the
following two coefficients:
j : the relative backward linkage coefficient in pollution. This shows the importance
that the returns and pollution of sector j have
in the total pollution that can be assigned to
the productive sector. It allows us to see the
polluting role played by one sector, in that it
incorporates the inputs of the other sectors
(that pollute) in its productive process.
j
1
n2
1
n vckj
n
P
7
vckj
j1
In an analogous manner, we can construct
the relative forward linkage coefficients (i)
that demonstrate the extent to which the total
returns of one sector vary in the face of an
increase in the final demand of all the sectors
in relation to the average behaviour of the
productive system.
i
1
n2
1
n icki
n
P
8
icki
i1
Following the idea of Rasmussen (1956),
Pajuelo (1980) establishes a classification,
using the forward and backward linkage
coefficients, adapted to atmospheric pollution (see Table I).
The key sectors are identified as those with
a marked polluting character as the demander of inputs and as the suppliers of resources to other sectors. In this sense, any
pollution-reducing activity in these key sectors that might affect their production could
have serious consequences for the whole
productive system.
However, it must be said that the use of
these indicators to determine key sectors has
been heavily criticised, in that they do not
provide information on their degree of
Table I
Classification using the forward and backward
linkages of coefficients
mj > 1
mj < 1
mj > 1
Key sectors
mj < 1
Sectors that
exhibit backward
linkage in the
pollution
Sector that exhibits
forward linkage
in the pollution
Other sectors
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
dispersion; that is to say, on the extent to
which the pollution generated by way of
intersectorial relationships is concentrated
in a reduced set of sectors or whether it
depends on the operation of almost all the
sectors of the economy. In order to examine
more deeply the information offered by the
relative backward and forward linkage coefficients, we can construct the backward and
forward linkage variation coefficients in
returns and pollution. These indicators try to
detect the level of pollution dispersion that
exists in the relationships of the economy. On
the basis of the indicators proposed by
AlcaÂntara (1995), we can, for any pollutant k
and for any sector j, construct the following
indicators:
Backward linkage variation coefficient in
type k pollution:
v
u
!2
uP
n
P
un
cki ij ÿ n1
cki ij
u
i1
1 u
ui1
CVAckj
u
n
P
n ÿ1t
1
cki ij
n
i1
Forward linkage variation coefficient in type
k pollution:
v
u
!2
u
n
n
P
P
u
1
c
c
ÿ
u ki
ij
ij
n ki
j1
j1
1 u
u
CVIcki
u
n
P
n ÿ 1t
1
ij
n cki
the economic sectors. The trade-offs approach and, on this basis, the extension of the
concept of elasticity, allows us to examine
these types of relationships more deeply.
The way to construct the trade-offs indicators between environmental and economic
variables can be found in AlcaÂntara (1995).
Having obtaining the valuation in k-type
pollution, we can similarly construct the
valuations in terms of income and employment
^ where v is the vector of
VV v I ÿ Aÿ1 y
direct unitary coefficients of added value,
^ where l is the vector of
and VL l I ÿ Aÿ1 y
direct unitary coefficients of labour. We can
also obtain the income-pollution and employment-pollution trade-offs as follows:
TOv ; ck VCk ÿ1 VV
TOL ; ck VCk ÿ 1 VL
These indicators are ranked from the largest
to the smallest. Sectors with both indicators
lower than the average are sectors where
reductions in pollution have a smaller effect
on the economic variables of income and
employment than in other sectors, which
means that the ``cost'', in terms of these
economic variables of choosing these sectors
for the reduction of pollution will have a
lower impact than in the other sectors of the
economy.
i1
Given the actual construction of the indicators, the backward linkage variation coefficients take high values when the return or
the pollution incorporated by a sector comes
from purchasing relationships with a reduced number of sectors. Here we are talking
of sectors whose polluting character is determined by their direct or indirect dependence on the purchases made from a small
number of highly polluting sectors within the
economy.
Similarly forward linkage variation coefficients take high values when the ``sales'' of
pollution made directly or indirectly by a
sector are concentrated in a reduced group of
sectors.
Trade-offs between pollution and economic
variables
In the search for better pollution indicators
that might assist us in determining which
are the key sectors of the economy with
respect to any type of pollution, Karunaratne
(1989) has proposed the trade-offs between
energy and economic variables. The idea is
that any measure for the reduction of pollution must take into account the repercussions in terms of income or employment in
Application to the Ebro Valley
Having considered the theoretical basis, let
us now turn to the results that we have
obtained for the Ebro Valley. Although our
analysis has been done to the four economic
structures of Lerida-Tarragona, AragoÂn, Navarra and La Rioja, given the amount and
complexity of the results, we only present
now the results for the aggregate table of the
Valley.
We start with the determination of the key
sectors proposed by Rasmussen (1956). Having obtained the backward and forward
linkages for all the sectors[2], we calculate
the relative coefficients, which will allow us
to draw the role played by each sector in the
economy. Table II and Table III list these
relative coefficients for the Ebro Valley.
On the basis of these results we can obtain
a classification, for each type of pollutant, as
shown in Table IV.
Thus, from the aggregated results for the
Valley as a whole, we can note how all the
irrigated agricultural sectors have a greater
relative backward linkage capacity with
respect to the volume of returns, followed by
the food sectors, livestock and paper. With
[ 147 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
respect to BDO5 pollution, the relative backward linkage sectors are metals, dairy products and wine production and livestock.
With respect to TSS pollution, it is the
energy, non-metal, paper and food sectors
that display the greatest backward linkage,
with the other sectors displaying values that
are similar amongst themselves but significantly lower than these first sectors. Finally,
with respect to nitrates and phosphates the
most relevant sectors are livestock (the great
value in both pollutants), followed by a part
of the food sector (mainly meat production)
and dairy products. Note that both sectors
are very joined to livestock production. After
these sectors the agricultural sectors are the
ones with a great relative backward linkage.
Similarly we can comment on the main
results obtained in the analysis of the relative forward linkages. With respect to the
volume of returns, the irrigated agricultural
sectors have the greatest values, followed by
chemicals and paper. In BDO5 the main
relative forward linkage is for dairy products
and juices, metals and livestock. In TSS the
classification is different: energy, non-metals, paper and chemicals are the sectors with
the greatest coefficient, although the food
industry also obtains great values. In nitrates
and phosphates only the livestock sector is
Table II
Relative backward linkages ± the Ebro Valley
Sectors
Cereal crops
under irrigation
Industrial crops
Vegetables
Fruits
Other crops
under irrigation
Dry land crops
and others
Livestock
Energy
Metals
Non-metals
Chemicals
Motor vehicles
Dairy products
and juices
Wine production
Other foods
Manufacturing
Paper
Construction
Hotels and
catering
Health services
Other services
Average
[ 148 ]
Returns
BDO5
TSS
Nitrates
Phosphates
6.87
0.11
0.14
0.49
0.46
2.59
2.01
1.86
4.15
0.11
0.12
0.11
0.12
0.14
0.14
0.14
0.14
0.54
0.52
0.53
0.50
0.54
0.51
0.52
0.48
0.03
0.12
0.14
0.55
0.55
0.49
0.02
0.04
0.14
0.39
0.02
0.65
0.85
0.03
1.47
0.09
0.12
0.15
15.56
0.13
2.10
0.35
7.23
1.72
0.31
3.09
13.67
0.00
0.00
0.00
0.01
0.00
1.46
13.73
0.00
0.00
0.00
0.01
0.00
1.47
0.48
0.45
0.07
0.46
0.03
0.17
1.16
0.20
0.12
0.20
0.10
0.19
1.60
0.26
0.65
1.42
0.81
0.31
0.10
2.25
0.05
0.11
0.00
0.19
0.10
2.26
0.05
0.11
0.00
0.19
0.04
0.04
1.00
0.03
0.02
1.00
0.07
0.10
1.00
0.01
0.01
1.00
0.01
0.01
1.00
representative in terms of relative forward
linkage.
If we pay attention to the classification
derived from the relative forward and backward linkages, the results for the Valley can
be compiled as follows: With respect to the
volume of returns, the key sectors are cereal
crops under irrigation, industrial crops,
fruits and other crops under irrigation, if we
consider all the sectors of the economy. If we
only consider the industrial sectors, the
chemical industry and paper sectors are the
most relevant in terms of relative forward
and backward linkage, although the food
sectors also have an important backward
linkage (due to the strong relationship with
agricultural production). Note that in LeridaTarragona and La Rioja, the chemical sector
does not appear to be important from the
forward linkage point of view.
As regards BDO5 pollution, the metal
sector, and food production sectors appear to
be key and only the livestock sector exhibits
forward linkage. This sector pushes the BDO5
pollution in the food sectors mainly.
For TSS pollution, the key sectors are
energy and non-metals, chemicals, agrifood
production and paper, although these results
are not common for all the regions.
Finally, with respect to the other pollutants, the livestock sector obtains the higher
values of these indicators in all the regions.
As a result, we can see that in all cases
there are three significant sectors, either as
key sectors or as sectors that exert backward
linkage over the others, thereby forcing
water pollution, namely the agriculture, livestock and agro-food sectors. Also note the
difference between sectors with respect to the
different types of pollutants. The importance
of the backward linkage in terms of BDO5 and
TSS (except for the livestock sector) is
concentrated fundamentally in the industrial
sectors (chemicals, paper, metal and motor
vehicles), whilst the responsibility for returns lies almost entirely with agriculture
and the sectors which depend on agricultural
production (agro-food, paper). The paper
sector stands out with respect to the volume
of returns. This is quite logical, given that
this sector receives inputs from the agricultural sector (a large-scale consumer and also
with important volumes of returns), whilst it
also supplies resources to almost all the
sectors of the economy. The chemical industry also appears as an important sector from
the point of view of water pollution,
especially in AragoÂn and Navarra. Both in
the different regions and in the Valley as a
whole, there appear to be no clear forward
linkages. In nitrates and phosphates the
livestock sector is the only one relevant.
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
Again we can note that the key sectors are
identified as those with a marked polluting
character as the demanders of inputs and as
the suppliers of resources to other sectors. In
this sense, any pollution-reducing activity in
these key sectors that might affect their
production could have serious consequences
for the whole productive system.
The next question we must raise concerns
the character of the ``sales and purchases'' of
pollution made by the sectors, that is to say,
the pollution incorporated by those sectors
with backward linkage potential, or introduced by those with forward linkage potential in their direct or indirect sales: does this
result from transactions with the majority of
the sectors in the economy or with just a
reduced number of these sectors? Table V
shows the coefficients of variation in the
backward and forward linkages:
The Table contains the results obtained for
the Ebro Valley. Note that it is the sectors
which reflect agriculture under irrigation
and basic industry, especially non-metals,
chemicals and paper, that obtain the highest
values for the variation coefficient in backward linkage in terms of returns. This means
that the returns which these sectors incorporate into their production have their origin
Table III
Relative forward linkages ± the Ebro Valley
Sectors
Cereal crops
under irrigation
Industrial crops
Vegetables
Fruits
Other crops
under irrigation
Dry land crops
and others
Livestock
Energy
Metals
Non-metals
Chemicals
Motor vehicles
Dairy products
and juices
Wine production
Other foods
Manufacturing
Paper
Construction
Hotels and
catering
Health services
Other services
Average
Returns
BDO5
TSS
Nitrates
Phosphates
7.57
0.00
0.00
0.04
0.00
2.73
2.20
1.97
4.19
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.01
0.01
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.39
0.02
0.06
0.14
0.45
0.01
0.19
1.23
0.00
2.40
0.03
0.12
0.07
15.62
0.00
3.18
0.44
8.32
1.81
0.20
2.99
20.91
0.00
0.00
0.00
0.00
0.00
0.00
21.00
0.00
0.00
0.00
0.00
0.00
0.00
0.24
0.06
0.08
0.48
0.01
0.10
1.16
0.03
0.14
0.19
0.00
0.01
1.43
0.22
0.81
1.45
0.00
0.14
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.03
0.07
1.00
0.00
0.00
1.00
0.00
0.00
1.00
0.00
0.00
1.00
0.00
0.00
1.00
in the inputs which are directly or indirectly
incorporated from a reduced number of
sectors. In general, the agricultural sectors
incorporate returns in their own self-consumption, whilst the industrial sectors produce their returns partly through their own
direct contribution and partly from the
inputs coming from other sectors (in the case
of paper, probably from the agricultural
sectors). With respect to the variation in the
forward linkages, we can note that it is the
cereal crops under irrigation sector which
obtains the highest value of the coefficient,
showing that the returns directly or indirectly incorporated by this sector (which, in
relation to the average for the economy, is
high), are concentrated in a very limited
number of sectors (mainly self-consumption
and the agrifood industries). Both with
respect to forward and backward linkages, it
is the final industry and services sectors that
maintain the most homogenous relations
with all the sectors, in such a way that the
returns they incorporate into their own
production or into the production of other
sectors is made via the transactions they
carry out with the majority of the sectors in
the economy.
Turning now to BDO5 pollution, it is
interesting to note that the highest values of
the variation coefficients in backward linkages are obtained in a number of groups:
first, in the livestock and energy sectors
(essentially by way of their direct pollution);
secondly, in the agrifood industry sectors
(fundamentally because of their purchasing
relationships with the livestock sector); and
finally, in the paper and construction sectors,
with both of these being large purchasers of
inputs from the energy sectors. Thus, we can
say that a significant incorporation of BDO5
results from the relationships that some
primary sectors (large direct polluters), specifically energy and livestock, maintain with
their most important purchasers of inputs.
The variation coefficients in backward linkages in BDO5 show that it is the energy
sector which directly or indirectly incorporates a large volume of pollution to a small
group of sectors. The observations made with
respect to BDO5 can be repeated in very
similar terms with respect to TSS pollution,
save for the livestock sector, which looses
importance in this regard. In the case of both
BDO5 and TSS, it is the services sectors
which show the greatest dispersion, followed
by the final industry sectors (motor vehicles,
manufacturing), confirming that the polluting character of these sectors is due to the
numerous contributions of pollution they
receive via the inputs which they incorporate
from almost all the sectors of the economy.
[ 149 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
Thus, any measure aimed at reducing BDO5
or TSS pollution must, in principle, be
directed towards controlling the pollution
that is emitted by the sectors with the highest
concentration of forward or backward linkages, given that in this way, and by acting
against just a few sectors, the majority of the
pollution in the system will be controlled.
As regards nitrates and phosphates, all the
sectors obtain similar values for the variation coefficients in backward linkages. As we
can see from the earlier tables, this type of
pollution is, in direct terms, concentrated in
the agriculture and livestock sectors, particularly in the second of these, and to some
extent in the hotels and catering sector.
Thus, the pollution which is directly and
indirectly generated by each sector of the
economy will only be due to the relationships
maintained with these sectors. In the variation in forward linkages, we can see that it is
the livestock sector which incorporates its
pollution in the sales that it makes to a larger
number of sectors, whilst agriculture concentrates its pollution to a greater extent
(which is logical, given that the livestock
sector sells inputs not only to the agrifood
industry ± dairy and meat products ±, but
also to the hotel and catering and to the
manufacturing sectors).
In summary, we can say that for the area
under study, and save for the differences
which exist for each type of pollutant, the
pollution that is directly and indirectly
generated in the system is not, in general
terms, spread homogeneously. Rather, it is
Table IV
Classification of pollutants
Pollutant
Key sectors
Returns
BDO5
Nitrates
Irrigated crops
Metals, dairy
products, wine,
Energy, nonmetals,
chemicals,
dairy
products,
wine,
paper
Livestock
Phosphates
Livestock
TSS
[ 150 ]
Sectors that
exhibit backward
linkage in the
pollution
Sectors that
exhibit forward
linkage in the
pollution
None
None
None
Livestock
None
None
Dairy products
and juices,
other foods
Dairy products
and juices,
other foods
None
None
concentrated in the sales made by a limited
number of highly polluting sectors (fundamentally, livestock and primary industries)
to other final goods sectors (agrifood, paper,
construction) which make the biggest purchases of inputs. The results describe an
economy that is not particularly integrated, a
panorama which, in terms of pollution, can
be translated into very specific groups of
large-scale polluters and pollution receivers.
Following this line, the last stage is to
analyse the trade-offs between economic and
environmental variables. We have already
said that any environmental planning measures which seek to control water pollution
must be directed towards these polluting
``groups'', given that a large part of the
pollution associated with the system would
thereby be controlled. At this point, we
should recall that any pollution control
measure which has an impact on the production of the sectors must take economic
aspects into account, specifically the total
contribution made by these sectors to the
income and employment of the area in
question. It is necessary to place the returns
or the total pollution incorporated into a
sector in relation with the total income and
employment it generates. AlcaÂntara (1994)
interprets the trade-offs with an opportunity
cost measurement, in terms of income or
employment, which introduces anti-pollution
measures that might have an impact on the
demand and production of the sectors. The
results derived from this trade-off technique
are set out in Table VI.
The lowest values of the trade-offs are
obtained in almost all cases for the agriculture, livestock and agrifood industry, save in
the case of TSS, a form of pollution where the
lowest trade-offs correspond to non-metals,
followed by dairy products, chemicals and
paper. In all cases the highest values are
obtained in the services sectors, generally
followed by motor vehicles, energy and construction. The results obtained for agriculture and livestock are easily understood:
these are sectors in which the income or
employment that is generated per unit of
pollution directly or indirectly incorporated
into production is lower than in the other
sectors or, put another way, the ``profitability'' obtained from the incorporation of
pollution is lower. At this point, it is interesting to focus our attention on the results
obtained for the food industry. Here, the low
trade-off obtained in almost all the cases
shows us that we are considering a sector in
which the pollution that is incorporated
directly and, especially, indirectly (forcing
the production of polluting outputs in the
agricultural or livestock sectors), does not
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
``compensate'' for the generation of income or
employment in this sector. In these sectors,
any activities designed to reduce pollution
have a low ``opportunity cost'', that is to say,
they are sectors in which a reduction in
pollution (which could be associated with a
reduction in the demand or production of
these sectors), will have lower effects over
the income and employment of the economy
than any activities taken in other sectors. In
the case of BDO5 or TSS pollution, we can see
how the livestock and food sectors are united
with the metals, non-metals, chemicals and
paper sectors. As regards pollution caused by
nitrates and phosphates, the livestock sector
and the related food sector (dairy products
and meats) obtain the lowest values.
If we now relate these results with those
obtained earlier, we get a clearer idea of the
distribution of returns and pollution in our
economy and can say that, when studying the
water pollution in the Ebro Valley, we
encounter a number of well defined returns
and pollution ``distribution groups''. In the
case of returns, the key sectors are those
which comprise agriculture under irrigation.
These sectors have a very high volume of
direct returns and their self-consumption of
inputs is important. The returns generated
by the agriculture sector are either returned
directly or incorporated into the production
of a reduced number of sectors (some into
livestock, some into hotel and catering, some
into the paper sector). In the paper sector,
although the direct volume of returns is
important, its character is marked by its
dependence on the raw materials from the
agriculture sector. The income and employment that this sector generates is important,
so it would be reasonable in this case to
control the returns via a control over agricultural activities (for example, by introducing measures for the reuse of irrigation
water, the improvement of irrigation systems, etc.). In the case of BDO5 we can
identify a highly polluting sector, namely
livestock, which introduces pollution into the
system through the sales that it makes to the
food and hotel and catering sectors, etc.
Together with this sector, primary industry
(metals, non-metals) also shows important
forward and backward linkages. Therefore,
in this case we are faced with two large
blocks of polluting sectors: those related most
directly with livestock and those related with
basic industry (metals, non-metals, chemicals). The study of the trade-offs shows us
that the profitability obtained from pollution
Table V
Variation coefficients in the Ebro Valley
Sector*
Returns
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Average
4.5654
4.5207
4.5081
4.4979
4.5596
1.6948
2.5524
2.8211
3.5375
4.1507
4.4171
2.0851
1.8840
2.3388
2.1428
3.5411
4.3346
2.1102
2.4812
3.1447
3.3124
3.2953
Variation coefficients in backward linkages
TSS
Nitrates
Phosphates
BD05
3.4173
3.2675
3.3909
3.2637
3.4431
3.3181
4.3631
4.1795
4.5650
3.1917
3.6919
3.0147
4.5424
4.4010
3.0328
3.6182
4.0326
4.2117
2.9039
2.7464
2.9971
3.5996
2.6387
2.4042
2.3748
2.4690
2.5064
2.3049
2.0768
4.4650
3.5235
4.4963
4.0678
2.9231
4.3820
3.9187
2.7454
3.8996
4.2545
4.1470
2.0441
2.2889
2.1895
3.1486
4.2636
4.5069
4.5214
4.5295
4.3711
4.5183
4.5822
4.5591
4.5701
4.5661
4.5715
4.5735
4.5744
4.5201
4.5787
4.5808
4.5792
4.5664
4.5441
4.5758
4.5688
4.5344
1.0239
1.0245
1.0246
1.0246
1.0242
1.0245
1.0247
1.0234
1.0240
1.0238
1.0243
1.0242
1.0247
1.0245
1.0247
1.0246
1.0247
1.0238
1.0224
1.0246
1.0241
1.0242
Returns
4.1283
0.0389
0.0425
0.0456
0.0264
0.0000
0.0855
0.4214
0.2160
0.1614
0.0908
0.7320
0.1504
0.1337
0.2246
0.1868
0.0919
0.5069
0.2027
0.4187
0.1418
0.4023
Variation coefficients in forward linkages
BD05
TSS
Nitrates
Phosphates
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0249
8.9242
0.0171
0.1788
0.0913
0.1333
0.0086
0.0311
0.1651
0.0751
0.0745
0.0000
0.3728
0.0000
0.0000
0.8414
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0118
0.0308
0.0085
0.0181
0.0596
0.0153
0.0218
0.0466
0.0241
0.0210
0.0000
0.0681
0.0000
0.0000
0.0296
4.1283
0.2707
0.2981
0.3264
0.1766
0.2581
0.0047
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.6959
0.0000
0.0000
0.7699
0.9231
0.7681
0.8458
0.9264
0.5013
0.7325
0.0014
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.3504
0.0000
0.0000
0.6311
Notes:
* 1 = cereal crops under irrigation; 2 = industrial crops; 3 = vegetables; 4 = fruits; 5 = other crops under irrigation; 6 = dry land crops and others;
7 = livestock; 8 = energy; 9 = metals; 10 = non-metals; 11 = chemicals; 12 = motor vehicles; 13 = dairy products and juices; 14 = wine
production; 15 = other foods; 16 = manufacturing; 17 = paper; 18 = construction; 19 = hotels and catering; 20 = health services; 21 = other
services
[ 151 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
in industry is higher than that obtained in
the livestock-food complex. Thus, this group
combines its high polluting potential with
the lower costs, in terms of income and
employment, of acting against it by way of
measures for the reduction of pollution. This
picture changes somewhat when we consider
TSS pollution. In this type of typically
industrial pollution, it is the energy, nonmetals and chemicals sectors that now play
the key role, although the food industry also
takes relevant values. Finally, as regards
nitrates and phosphates pollution, it is the
livestock sector which appears as key according to the Rasmussen classification,
whilst the food industry sectors show a
marked backward linkage potential in pollution. Once again, it is the livestock-food
complex which concentrates the highest
productive incorporation of pollution, with it
also showing the lowest trade-offs.
Conclusions
In this work, we have applied some of the
tools made available by the input-output
methodology to the study of water pollution
in the Ebro Valley. We have offered no more
than a limited analysis, one which only
purports to demonstrate some of the possibilities that can be found in the study of the
relationships between the economy and the
environment. Notwithstanding this qualification, our initial analysis allows us to draw
the following conclusions.
The proposed analysis demonstrates the
possibility of using the traditional economic
models for a more in-depth study of water
pollution associated with the economic
structure of a region. In this paper we have
used two methods to approximate the relationships between economic and environmental variables. The first is an application
of the forward and backward linkage coefficients proposed by Rasmussen (1956) and
adapted by Pajuelo (1980), Proops et al. (1993)
and AlcaÂntara (1995) to the case of atmospheric pollution, whilst the second is an
approximation of the so-called trade-offs
relationships between economic and water
pollution variables. This paper represents a
first attempt to construct both types of
indicators for the study of the water pollution
Table VI
Trade-offs in the Ebro Valley
Sectors*
Returns
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Average
0.09
0.25
0.31
0.34
0.10
18.93
0.95
47.67
10.04
4.75
1.19
18.86
0.78
1.20
0.96
8.13
1.04
18.93
3.92
21.92
21.23
8.65
Trade-offs added value-pollutiona
BD05
TSS
Nitrates
1.39
1.50
1.36
1.52
1.33
1.42
0.15
7.20
0.08
1.91
1.00
0.64
0.01
0.13
0.58
1.32
0.63
1.70
0.91
7.57
10.95
2.06
13.38
14.50
14.16
14.30
13.65
14.65
11.24
1.17
4.06
0.28
0.87
3.61
0.54
1.17
5.41
2.96
1.07
2.57
6.66
37.58
26.50
9.06
0.20
0.19
0.20
0.20
0.20
0.19
0.01
74.39
25.00
36.19
10.73
16.22
0.06
0.94
0.03
1.95
0.74
29.68
0.57
8.84
22.68
10.91
Phosphates
Returns
0.35
0.32
0.33
0.33
0.34
0.32
0.01
121.94
40.92
59.26
17.57
26.54
0.09
1.56
0.05
3.18
1.21
48.61
0.93
14.45
37.13
17.88
0.03
0.08
0.11
0.12
0.04
5.96
0.51
5.21
2.50
1.00
0.30
3.69
0.26
0.28
0.35
2.98
0.25
6.46
1.31
6.38
7.25
2.15
Trade-offs employment-pollutionb
BD05
TSS
Nitrates
0.54
0.49
0.46
0.52
0.49
0.45
0.08
0.79
0.02
0.40
0.25
0.12
0.00
0.03
0.21
0.49
0.15
0.58
0.30
2.20
3.74
0.59
5.23
4.75
4.81
4.86
5.05
4.61
6.08
0.13
1.01
0.06
0.22
0.71
0.18
0.27
1.96
1.09
0.26
0.88
2.22
10.94
9.05
3.07
0.08
0.06
0.07
0.07
0.07
0.06
0.00
8.13
6.23
7.61
2.66
3.17
0.02
0.22
0.01
0.71
0.18
10.13
0.19
2.57
7.74
2.38
Phosphates
0.14
0.11
0.11
0.11
0.13
0.10
0.00
13.33
10.19
12.47
4.36
5.19
0.03
0.36
0.02
1.17
0.29
16.58
0.31
4.21
12.68
3.90
Notes:
a
Returns in thousands of pesetas/cubic metre, pollution in thousands of pesetas/kg of pollutant
b
Returns in thousands of jobs/cubic hectometre, pollution in thousands of jobs/kg of pollutant
* 1 = cereal crops under irrigation; 2 = industrial crops; 3 = vegetables; 4 = fruits; 5 = other crops under irrigation; 6 = dry land crops and others;
7 = livestock; 8 = energy; 9 = metals; 10 = non-metals; 11 = chemicals; 12 = motor vehicles; 13 = dairy products and juices; 14 = wine
production; 15 = other foods; 16 = manufacturing; 17 = paper; 18 = construction; 19 = hotels and catering; 20 = health services; 21 = other
services
[ 152 ]
Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure
and water pollution in the Ebro
Valley (Spain)
Environmental Management
and Health
10/3 [1999] 143±154
associated with a specific economic system,
in our case, the Ebro Valley in Spain, with
attention being given to a series of pollutants
that are characteristic of the agricultural,
livestock and industrial activities carried out
in that system. The results obtained from the
first analysis show that, in general terms, it
is the agriculture, livestock and agrifood
sectors which directly and indirectly emit
the highest volumes of returns and of nitrate
and phosphates pollution and for which the
``opportunity costs'' are lower in terms of
added value and employment. This situation
extends to the paper sector (also in terms of
volume of returns) and to the metals and nonmetals