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 Sa
Ânchez-Cho
Âliz
Department of Economic Analysis, Faculty of Economics and Business Studies,
University of Zaragoza, Spain
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 en-vironmental 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 re-sponsibilities for environmental pollution, there are few studies that examine precisely which economic sectors have a highly pol-luting 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 chal-lenge 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 eco-nomic activities and the environmental con-ditions 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 agricul-tural 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 in-dustrial areas and the ever-increasing com-plexity 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 responsibil-ity 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 determi-nation of ``key sectors'', but introducing some qualifications with respect to the indicators obtained; and with this analysis being com-pleted with a brief description of the respon-sibilities of the different components of the economic demand of the system; secondly, by considering the relationships between eco-nomic 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-eco-nomic 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 antece-dents, the initial assumptions, the criteria for the aggregation of the sectors and the homo-genisation of the tables and the character-isation of water pollution. In Section 3 we adopt a theoretical approach towards ob-taining the indicators derived from the input-output model. In a first analysis we refer to the characterisation of key sectors following Rasmussen, and subsequently to the obtain-ing of the trade-offs. In Section 4 we comment on the results obtained from an application of this methodology to the Ebro Valley. Section Environmental Management
and Health
10/3 [1999] 143±154
#MCB University Press [ISSN 0956-6163]
Keywords
Input/output analysis, Structure, Pollution, Economics
Abstract
5 closes the paper with a review of the most relevant conclusions.
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 instru-ments for estimation and economic forecast-ing. The explanatory power of these models has also been taken advantage of in environ-mental 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 tradi-tional 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 environmen-tal 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 be-tween 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 concen-trates 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 pol-lution 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 re-ceiver of residues). This author's work sup-posed 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 en-vironmental sectors themselves. Neverthe-less, he himself recognised the impossibility of obtaining data on these latter relation-ships, and in a first analysis, the flows are recognised in monetary terms. Another im-portant work is that of Victor (1972), who proposed the construction of satellite ac-counts which, when united with the
eco-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 princi-pally 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-out-put 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 envir-onmental 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 pollut-ing 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 char-acterised by these four indicators, they do nevertheless offer a first synthesis of the agricultural and industrial pollution (sup-posing 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 pur-poses.
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
The multisectorial descriptions have been obtained from the regional tables corre-sponding to each area. The study has been carried out for the following 21 representa-tive sectors: cereals under irrigation, indus-trial 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 ser-vices. 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 em-ployment and production statistics, etc. Si-milarly, 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 theAgrarian 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 CatalunÄ a (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 func-tion 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 un-polluted 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, whereXjrepresents the effective production of sector j,Yjthe destination of theXjto the final demand andXithe use that sector j makes of the production of sector i , we have for any sector j:
XjXj1Xj2. . .XjnYj
Accepting the hypothesis that the proportion of factors used by each sector is fixed, the technical coefficientsaijcan be defined as:
aijXij=Xj
For all the sectors of the economy and in a matrix form, the system can be expressed as:
XAXY 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
ckjrepresents 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:
ECXCAXCY 2 where E is the matrix of total returns and pollutions of the economy.
Expression (1) can also be written as:
X IÿAÿ1Y 3 where IÿAÿ1is the Leontief inverse ma-trix. The total by columns of theijelements of the inverse provide us, for each sector, with the direct and indirect input require-ments when the final demand of each sector increases by one unit.
If we pre-multiply by C, we obtain:
VCCX C IÿAÿ1Y 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.
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
Furthermore, in unitary terms we can propose the following indicators:
vcC IÿAÿ1
is the matrix of ``pollution values'' associated with the backward linkages or unitary coef-ficients 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 back-ward 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 inter-pretation of these coefficients has to be made in terms of ``values'' (the Marx value of labour, or water value (SaÂnchez-ChoÂlizet al., 1994)). The interpretation of these co-effi-cients as values opens an avenue for us to compare these with the monetary values of specific types of economic activities. Simi-larly, these coefficients are interpreted by Pasinetti (1977) as vertical integration coeffi-cients.
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 ma-trix, 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 produc-tion 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:
icki
Xn
j1
ij 6
which is the returns or pollution generated by the ithsector (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 differ-ent regions on the part of particular sectors.
On the basis of such indicators, Pajuelo (1980) adapts those constructed by Rasmus-sen (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 coeffi-cient 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
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.
Following the idea of Rasmussen (1956), Pajuelo (1980) establishes a classification, using the forward and backward linkage coefficients, adapted to atmospheric pollu-tion (see Table I).
The key sectors are identified as those with a marked polluting character as the deman-der of inputs and as the suppliers of re-sources to other sectors. In this sense, any pollution-reducing activity in these key sec-tors 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 Sector that exhibits
forward linkage
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
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 coef-ficients, 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:
Forward linkage variation coefficient in type k pollution:
Given the actual construction of the indica-tors, the backward linkage variation coeffi-cients take high values when the return or the pollution incorporated by a sector comes from purchasing relationships with a re-duced number of sectors. Here we are talking of sectors whose polluting character is de-termined by their direct or indirect depen-dence on the purchases made from a small number of highly polluting sectors within the economy.
Similarly forward linkage variation coeffi-cients 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 pollu-tion must take into account the repercus-sions in terms of income or employment in
the economic sectors. The trade-offs ap-proach 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 indica-tors 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 employ-ment
VV v IÿAÿ1^ywhere v is the vector of direct unitary coefficients of added value, andVLl IÿAÿ1^ywhere l is the vector of direct unitary coefficients of labour. We can also obtain the income-pollution and em-ployment-pollution trade-offs as follows:
TOv;ckVCkÿ1VV TOL;ckVCkÿ1VL
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.
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, Na-varra 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). Hav-ing 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 Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
respect to BDO5pollution, the relative
back-ward linkage sectors are metals, dairy pro-ducts 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 signifi-cantly 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 rela-tive forward linkages. With respect to the volume of returns, the irrigated agricultural sectors have the greatest values, followed by chemicals and paper. In BDO5the main
relative forward linkage is for dairy products and juices, metals and livestock. In TSS the classification is different: energy, non-me-tals, 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
representative in terms of relative forward linkage.
If we pay attention to the classification derived from the relative forward and back-ward 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 Lerida-Tarragona and La Rioja, the chemical sector does not appear to be important from the forward linkage point of view.
As regards BDO5pollution, 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 pollu-tants, 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, live-stock 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 BDO5and
TSS (except for the livestock sector) is concentrated fundamentally in the industrial sectors (chemicals, paper, metal and motor vehicles), whilst the responsibility for re-turns 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 agricul-tural 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 indus-try 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. Table II
Relative backward linkages ± the Ebro Valley
Sectors Returns BDO5 TSS Nitrates Phosphates
Cereal crops under irrigation
6.87 0.11 0.14 0.49 0.46
Industrial crops 2.59 0.11 0.14 0.54 0.54
Vegetables 2.01 0.12 0.14 0.52 0.51
Fruits 1.86 0.11 0.14 0.53 0.52
Other crops under irrigation
4.15 0.12 0.14 0.50 0.48
Dry land crops and others
0.03 0.12 0.14 0.55 0.55
Livestock 0.49 0.85 0.13 13.67 13.73
Energy 0.02 0.03 2.10 0.00 0.00
Metals 0.04 1.47 0.35 0.00 0.00
Non-metals 0.14 0.09 7.23 0.00 0.00
Chemicals 0.39 0.12 1.72 0.01 0.01
Motor vehicles 0.02 0.15 0.31 0.00 0.00
Dairy products and juices
0.65 15.56 3.09 1.46 1.47
Wine production 0.48 1.16 1.60 0.10 0.10
Other foods 0.45 0.20 0.26 2.25 2.26
Manufacturing 0.07 0.12 0.65 0.05 0.05
Paper 0.46 0.20 1.42 0.11 0.11
Construction 0.03 0.10 0.81 0.00 0.00
Hotels and catering
0.17 0.19 0.31 0.19 0.19
Health services 0.04 0.03 0.07 0.01 0.01
Other services 0.04 0.02 0.10 0.01 0.01
Average 1.00 1.00 1.00 1.00 1.00
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
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 intro-duced by those with forward linkage poten-tial 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 back-ward linkage in terms of returns. This means that the returns which these sectors incor-porate into their production have their origin
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-con-sumption, whilst the industrial sectors pro-duce 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 indir-ectly 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 BDO5pollution, it is
interesting to note that the highest values of the variation coefficients in backward lin-kages 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), spe-cifically energy and livestock, maintain with their most important purchasers of inputs. The variation coefficients in backward lin-kages in BDO5show that it is the energy
sector which directly or indirectly incorpo-rates a large volume of pollution to a small group of sectors. The observations made with respect to BDO5can 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 BDO5and TSS, it is the services sectors
which show the greatest dispersion, followed by the final industry sectors (motor vehicles, manufacturing), confirming that the pollut-ing 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. Table III
Relative forward linkages ± the Ebro Valley
Sectors Returns BDO5 TSS Nitrates Phosphates
Cereal crops under irrigation
7.57 0.00 0.00 0.04 0.00
Industrial crops 2.73 0.00 0.00 0.01 0.00
Vegetables 2.20 0.00 0.00 0.01 0.00
Fruits 1.97 0.00 0.00 0.01 0.00
Other crops under irrigation
4.19 0.00 0.00 0.02 0.00
Dry land crops and others
0.00 0.00 0.00 0.01 0.00
Livestock 0.39 1.23 0.00 20.91 21.00
Energy 0.02 0.00 3.18 0.00 0.00
Metals 0.06 2.40 0.44 0.00 0.00
Non-metals 0.14 0.03 8.32 0.00 0.00
Chemicals 0.45 0.12 1.81 0.00 0.00
Motor vehicles 0.01 0.07 0.20 0.00 0.00
Dairy products and juices
0.19 15.62 2.99 0.00 0.00
Wine production 0.24 1.16 1.43 0.00 0.00
Other foods 0.06 0.03 0.22 0.00 0.00
Manufacturing 0.08 0.14 0.81 0.00 0.00
Paper 0.48 0.19 1.45 0.00 0.00
Construction 0.01 0.00 0.00 0.00 0.00
Hotels and catering
0.10 0.01 0.14 0.00 0.00
Health services 0.03 0.00 0.00 0.00 0.00
Other services 0.07 0.00 0.00 0.00 0.00
Average 1.00 1.00 1.00 1.00 1.00
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
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 lin-kages, 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 varia-tion 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, parti-cularly 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 varia-tion 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 con-centrates 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
concentrated in the sales made by a limited number of highly polluting sectors (funda-mentally, livestock and primary industries) to other final goods sectors (agrifood, paper, construction) which make the biggest pur-chases 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 mea-sures 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 pro-duction 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 agricul-ture, 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 con-struction. The results obtained for agricul-ture 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 ``profitabil-ity'' obtained from the incorporation of pollution is lower. At this point, it is inter-esting 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 Table IV
Returns Irrigated crops None None
BDO5 Metals, dairy
products, wine,
Nitrates Livestock Dairy products
and juices, other foods
None
Phosphates Livestock Dairy products
and juices, other foods
None
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
``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 BDO5or 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 employ-ment that this sector generates is important, so it would be reasonable in this case to control the returns via a control over agri-cultural activities (for example, by introdu-cing measures for the reuse of irrigation water, the improvement of irrigation sys-tems, etc.). In the case of BDO5we 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, chemi-cals). The study of the trade-offs shows us that the profitability obtained from pollution
Table V
Variation coefficients in the Ebro Valley
Variation coefficients in backward linkages Variation coefficients in forward linkages
Sector* Returns BD05 TSS Nitrates Phosphates Returns BD05 TSS Nitrates Phosphates
1 4.5654 3.4173 2.6387 4.2636 1.0239 4.1283 0.0000 0.0000 4.1283 0.9231
2 4.5207 3.2675 2.4042 4.5069 1.0245 0.0389 0.0000 0.0000 0.2707 0.7681
3 4.5081 3.3909 2.3748 4.5214 1.0246 0.0425 0.0000 0.0000 0.2981 0.8458
4 4.4979 3.2637 2.4690 4.5295 1.0246 0.0456 0.0000 0.0000 0.3264 0.9264
5 4.5596 3.4431 2.5064 4.3711 1.0242 0.0264 0.0000 0.0000 0.1766 0.5013
6 1.6948 3.3181 2.3049 4.5183 1.0245 0.0000 0.0000 0.0000 0.2581 0.7325
7 2.5524 4.3631 2.0768 4.5822 1.0247 0.0855 0.0249 0.0000 0.0047 0.0014
8 2.8211 4.1795 4.4650 4.5591 1.0234 0.4214 8.9242 0.0118 0.0000 0.0000
9 3.5375 4.5650 3.5235 4.5701 1.0240 0.2160 0.0171 0.0308 0.0000 0.0000
10 4.1507 3.1917 4.4963 4.5661 1.0238 0.1614 0.1788 0.0085 0.0000 0.0000
11 4.4171 3.6919 4.0678 4.5715 1.0243 0.0908 0.0913 0.0181 0.0000 0.0000
12 2.0851 3.0147 2.9231 4.5735 1.0242 0.7320 0.1333 0.0596 0.0000 0.0000
13 1.8840 4.5424 4.3820 4.5744 1.0247 0.1504 0.0086 0.0153 0.0000 0.0000
14 2.3388 4.4010 3.9187 4.5201 1.0245 0.1337 0.0311 0.0218 0.0000 0.0000
15 2.1428 3.0328 2.7454 4.5787 1.0247 0.2246 0.1651 0.0466 0.0000 0.0000
16 3.5411 3.6182 3.8996 4.5808 1.0246 0.1868 0.0751 0.0241 0.0000 0.0000
17 4.3346 4.0326 4.2545 4.5792 1.0247 0.0919 0.0745 0.0210 0.0000 0.0000
18 2.1102 4.2117 4.1470 4.5664 1.0238 0.5069 0.0000 0.0000 0.0000 0.0000
19 2.4812 2.9039 2.0441 4.5441 1.0224 0.2027 0.3728 0.0681 0.6959 0.3504
20 3.1447 2.7464 2.2889 4.5758 1.0246 0.4187 0.0000 0.0000 0.0000 0.0000
21 3.3124 2.9971 2.1895 4.5688 1.0241 0.1418 0.0000 0.0000 0.0000 0.0000
Average 3.2953 3.5996 3.1486 4.5344 1.0242 0.4023 0.8414 0.0296 0.7699 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
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
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, non-metals 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 ac-cording to the Rasmussen classification, whilst the food industry sectors show a marked backward linkage potential in pollu-tion. 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 possibi-lities that can be found in the study of the relationships between the economy and the environment. Notwithstanding this qualifi-cation, 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 rela-tionships between economic and environ-mental variables. The first is an application of the forward and backward linkage coeffi-cients proposed by Rasmussen (1956) and adapted by Pajuelo (1980), Proopset al. (1993) and AlcaÂntara (1995) to the case of atmo-spheric 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
Trade-offs added value-pollutiona Trade-offs employment-pollutionb
Sectors* Returns BD05 TSS Nitrates Phosphates Returns BD05 TSS Nitrates Phosphates
1 0.09 1.39 13.38 0.20 0.35 0.03 0.54 5.23 0.08 0.14
2 0.25 1.50 14.50 0.19 0.32 0.08 0.49 4.75 0.06 0.11
3 0.31 1.36 14.16 0.20 0.33 0.11 0.46 4.81 0.07 0.11
4 0.34 1.52 14.30 0.20 0.33 0.12 0.52 4.86 0.07 0.11
5 0.10 1.33 13.65 0.20 0.34 0.04 0.49 5.05 0.07 0.13
6 18.93 1.42 14.65 0.19 0.32 5.96 0.45 4.61 0.06 0.10
7 0.95 0.15 11.24 0.01 0.01 0.51 0.08 6.08 0.00 0.00
8 47.67 7.20 1.17 74.39 121.94 5.21 0.79 0.13 8.13 13.33
9 10.04 0.08 4.06 25.00 40.92 2.50 0.02 1.01 6.23 10.19
10 4.75 1.91 0.28 36.19 59.26 1.00 0.40 0.06 7.61 12.47
11 1.19 1.00 0.87 10.73 17.57 0.30 0.25 0.22 2.66 4.36
12 18.86 0.64 3.61 16.22 26.54 3.69 0.12 0.71 3.17 5.19
13 0.78 0.01 0.54 0.06 0.09 0.26 0.00 0.18 0.02 0.03
14 1.20 0.13 1.17 0.94 1.56 0.28 0.03 0.27 0.22 0.36
15 0.96 0.58 5.41 0.03 0.05 0.35 0.21 1.96 0.01 0.02
16 8.13 1.32 2.96 1.95 3.18 2.98 0.49 1.09 0.71 1.17
17 1.04 0.63 1.07 0.74 1.21 0.25 0.15 0.26 0.18 0.29
18 18.93 1.70 2.57 29.68 48.61 6.46 0.58 0.88 10.13 16.58
19 3.92 0.91 6.66 0.57 0.93 1.31 0.30 2.22 0.19 0.31
20 21.92 7.57 37.58 8.84 14.45 6.38 2.20 10.94 2.57 4.21
21 21.23 10.95 26.50 22.68 37.13 7.25 3.74 9.05 7.74 12.68
Average 8.65 2.06 9.06 10.91 17.88 2.15 0.59 3.07 2.38 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
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
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 non-metals sectors for BDO5pollution. For TSS
pollution the results show a basic industrial predominance, although the livestock sector conserves its position as one of the greatest polluters.
The study of the trade-offs leads us to conclude that for the livestock, agriculture under irrigation and agrifood sectors, the creation of income and employment per unit of return and pollution generated is mini-mised, showing that these sectors are parti-cularly relevant when considering sectorial measures for the reduction of pollution which take their economic repercussions into account. The impact for the economy as a whole of reducing the pollution of these sectors would take the form of a marked reduction in the pollution of the economy at a relatively low cost in terms of added value and employment.
Finally, an analysis of this type allows us to consider a more extended table which, in subsequent studies, will incorporate the purification of waste water and its economic costs associated with the input-output model. In this sense, Article 2 of EU Directive 91/ 271/EEC, of 21 May 1991, establishes that pollution in terms of BDO5of 60gr/day
corresponds to one equivalent inhabitant. Furthermore, both the BDO5and TSS
mea-sures are frequently used to determine the basic tariffs to be paid in urban centres for the purification of urban water.
Using this as a basis, and according to the characteristics of the returns and the pur-ifying conditions, we could determine the purification cost ratios, which would be a way of relating, at least to a certain extent, the purification costs with the returns di-rectly and indidi-rectly generated by the indus-tries. A whole new line of research would thereby by opened, in which we could look for new pollution indicators and for an extension of the traditional input-output tables in which the environmental sector could be reflected in a more coherent man-ner.
Notes
1 The Ebro Basin is bordered by the Pyr-inees mountains to the north, the Medi-terranean Sea to the east and by the Iberian Chain to the south-west. The Ebro Valley is formed by four regional econo-mies: AragoÂn, Navarra, La Rioja and a part of Catalonia (Lerida and Tarragona). 2 The indicators for the four regions and for a level of aggregation of 12 sectors can be seen in Duarte and ChoÂliz (1998).
References
AlcaÂntara, V. (1995), ``EconomõÂa y contaminacioÂn atmosfeÂrica: hacia un nuevo enfoque desde el anaÂlisis input-output'', unpublished doctoral thesis, Barcelona.
Bielsa, J. (1998), ``ModelizacioÂn de la gestioÂn del agua integrada en el territorio: magnitudes asociadas desde una perspectiva econoÂmica'', unpublished doctoral thesis.
Daly, H. (1980), in Freeman, W.H. (Ed.),
Economics, Ecology, Ethics: Essays Toward a Steady-state Economy, San Francisco, CA. Duarte, R. and Choliz, J. (1998), ``Analysis of water
pollution with input-output models'',Applied Sciences and the Environment, Engineering Series, WIT Press, Southampton.
Karunaratne, N.D. (Ed.) (1989),Australian Development Issues: an Input-Output Analysis, Gower, Aldershot.
Leontief, W. (1970), ``Environmental repercussion and the economic structure: an input-output approach'',Review of Economics and Statistics, Vol. 11, August.
MAPA (1993),Anuario de EstadõÂstica Agraria 1992, Ministry of Agriculture, Fishing and Food, Madrid.
Miller, R.E. and Blair, P.D. (1985),Input-output Analysis: Foundations and Extensions, Pre-ntice-Hall, Englewood Cliffs, NJ.
Pajuelo, A. (1980), ``Equilibrio general versus anaÂlisis parcial en el anaÂlisis input-output econoÂmico ambiental: Una aplicacioÂn al anaÂlisis de la contaminacioÂn atmosfeÂrica en EspanÄ a'',Revista del Instituto de Estudios EconoÂmicos, No. 3, Madrid.
Pasinetti, L. (1977),Aportaciones a la teorõÂa de la produccioÂn conjunta, FCE, MeÂxico.
Proops, J. Faber, M. and Wagenhals, G. (1993),
Reducing CO2Emissions: a Comparative Input-Output Study for Germany and the UK, Springer-Verlag, Berlin.
Proops, J. Faber, M. and Wagenhals, G. (1993),
Reducing CO2 Emissions: a Comparative Input-Output Study for Germany and the UK, Springer-Verlag, Berlin.
Rasmussen, P. (Ed.) (1956),Relaciones Intersectoriales, Aguilar, Madrid.
SaÂnchez-ChoÂliz, J. (1997),Efectos EconoÂmicos y Sociales de las Inversiones en Riego Contempladas en la Propuesta del Plan HidroloÂgico de la Cuenca del Ebro, Confed-eracioÂn HidrograÂfica del Ebro, Zaragoza. Rosa Duarte Pac and
Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
SaÂnchez-ChoÂliz, J. and Duarte, R. (1997), ``Efectos econoÂmicos del Plan HidroloÂgico de la Cuenca del Ebro. AnaÂlisis a traveÂs del incremento de valor anÄ adido'', working paper, Department of Economic Analysis, University of Zaragoza.
SaÂnchez-ChoÂliz, J., Arrojo, P. and Bielsa, J. (1994), ``Valores agua para Aragon'', paper presented at theIV Jornadas de EconomõÂa CrõÂtica, Valencia.
Stone, R. (1980),Process, Capacity and Control in an Input-Output System, Cambridge. Tabuenca, J.M. (Ed.) (1994),Curso Sobre Uso,
Ahorro y Calidad del Agua, DGA, Zaragoza, Spain.
Victor (Ed.) (1972),Pollution: Economy and Environment, Allen and Unwin, London. World Health Organization (1982),Rapid
Assessment of Sources of Air, Water and Land
Pollution, WHO offset publication No. 62, Geneva.
Further reading
Leontief, W. and Ford, D. (1972), ``Air pollution and the economic structure: empiricial re-sults of input-output computations'', in Brody and Carter (Eds),Input-Output Techniques, North Holland Publishing, Amsterdam. Pulido, A. and Fontela, E. (1993),AnaÂlisis
Input-Output. Modelos, Datos y Aplicaciones, PiraÂ-mide, Madrid.
SubdireccioÂn General de EstadõÂstica y AnaÂlisis Sectorial (1995),Anuario de EstadõÂstica Agraria, 1993, Ministry of Agriculture, Fish-ery and Food, Madrid.
Rosa Duarte Pac and Julio SaÂnchez-ChoÂliz
Regional productive structure and water pollution in the Ebro Valley (Spain)
Environmental Management and Health
