The German PIOT yields the following degrees:
degree of linearity:0.96,
degree of interdependence/circularity:0.08.
These measures are near their extreme values (maximum/minimum); that is, the degree of linearity is very high and, conversely, the degree of interdependence/circularity is very low.3To present these results in a more meaningful form, the triangularized PIOT is fil- tered and transformed into Boolean form. Its elements are set equal to 1, ifxij> xji, and equal to 0, otherwise. Table 10.4 shows the extremely linear organization of the produc- tion system; that is, when the activities are presented in the order E, C, M, I, H, P, the result is a complete triangular matrix. That means that the primary material input is trans- formed along this activity chain without any feedback circuits. Not even environmental protection services (activity 6) creates a feedback.
Table 10.4 Filtered triangularized PIOT
E C M I H P
2 4 1 3 5 6
2 E 1 1 1 1 1
4 C 0 1 1 1 1
1 M 0 0 1 1 1
3 I 0 0 0 1 1
5 H 0 0 0 0 1
6 P 0 0 0 0 0 0
This result, which is incompatible with the common idea of a recycling economy (at least in Germany), underlines the crude fact that the German economy is a typical throughput economy (see below).
CONCLUSION: DEFICIENCIES OF THE MATERIAL
input of our industrial system depends heavily on the extraction of non-renewable raw materials (fossil fuels). ‘In this sense, the industrial system of today resembles the earliest stage of biological evolution, when the most primitive living organisms obtained their energy from a stock of organic molecules accumulated during prebiotic times’ (ibid., p. 44). Here we have one of the origins of anthropogenic emissions.
Downstream Dominance
The second characteristic is that the metabolism of living organisms (cells) is executed by multi-step regenerative chemical reactions in an aqueous medium at ambient tempera- tures and pressures (Ayres 1989a, p. 39). In contrast to this capability, our industrial pro- duction system can be characterized as a throughput economy where the industrial processes are irreversible transformations. A low transformation intensity follows from this. It is low because the industrial processes differ from biological organisms in that they are not (yet) able to build complex molecules directly from elementary building blocks with relatively few intermediates, as, for example, by the citric acid cycle in each cell of a living organism (ibid., p. 43). Within multi-step regenerative chemical reactions, con- trolled by catalysts (enzymes), most process intermediates are regenerated internally within the cell (ibid., p. 39).
Generally, a cyclic organization of production processes is a necessary condition of self-reproducing systems. If, as in industrial production, such a cyclic organization is absent and the system is dominated by process chains, self-reproduction does not take place and intermediates are embodied in downstream products or immediately wasted.
Material Flush-out
‘The third salient characteristic differentiating the biosphere from the industrial synthe- sphere is that, although individual organisms do generate process wastes – primarily oxygen in the case of plants and carbon dioxide and urea in the case of animals – the bio- sphere as a whole is extremely efficient at recycling the elements essential to life.
Specialized organisms have evolved to capture nutrients in wastes (including dead organ- isms) and recycle them’ (Ayres 1989a, p. 41).
These ‘specialized decay organisms’ (destruents) constitute a very important part of the biosphere, mainly of the pedosphere (soil), where they interact within a complex transfor- mation network. For an analytical description see, for example, the input–output frame- work for the natural production system of Strassert (1993, 1997) where four main groups of soil organisms are represented as production/transformation activities which interlink two food chains, the saprophage and the biophage food chain, on two levels, the aerobic and the anaerobic soil level. This important transformation domain is the final comple- ment of the cyclic organization of the material flow as a provision–transformation–resti- tution cycle (ibid.) which connects the last with the first domain; that is, the restitution domain with the provision domain.
The question arises as to whether, in our industrial production system, there exists such a transformation domain, the function of which can be compared with the final part of the digestive tract (Daly 1995, p. xiii) of the ‘organisms biosphere’. The answer must be negative. Most of our industrial recycling activities are far too small in scale to achieve a
comparable functional importance. Possibly the evolution of a new domain of industrial destruents in form of the specialized ‘cracker’ technologies is yet to come.
Open Conceptual Problems
In an early phase of physical input–output accounting it is quite natural that a lot of con- ceptual questions are still under discussion. An important example, out of a number of ambiguities, is the water problem. On the one hand, if the data are available, a compre- hensive approach is preferred (the German case); that is, all water quantities are counted, including those directly related to a production process (process water) as well as those indirectly related (cooling or irrigation water). However, it can be argued that ‘through- put’ water has a minimal environmental impact (except where it is very scarce) and that data in many countries are poor. In this case a more restrictive approach, in principle ori- ented to process water, is suggested (Ayres 2000; Gråvgard 1998; Nebbia 1999).
The general problem an accounting scheme should (ideally) avoid is that the overall total of all materials is dominated by the quantities of water. This refers not only to raw materials and residuals, but also to products, which also include water sold to households by water supply enterprises. So when, as in the German case, roughly two-thirds of the total quantity of products of the economy in tons is domestic water and the overall content of water in gross output is about 92 per cent, a PIOT is in danger of presenting only a more or less impure water account.
From this point of view, several authors (Ayres et al.2000; Gråvgard 1998; Nebbia 1999) propose a restrictive convention; namely that water participating in an economic process onlyas a passive carrier of heat or a diluent of waste should not be counted. On the other hand, water that participates actively in a chemical or biological process must be counted on both sides: that is, both as an input and as an output.4
Gråvgard proposes that the input of water be limited to the quantity of water supplied to (embodied in) products in the manufacturing industry, and which therefore leaves the industry again, together with the goods produced. Water supplied to products in agricul- ture, horticulture, forestry and fishery is implicitly included when calculating biomass weight. Additional water consumption, that is the water which evaporates on the output side or which the sectors discharge to the waste system and so on is not included (Gråvgard,1998, p. 9).
Nebbia (1999) too wants not to consider the water flow through the economic system, but only:
a. the amounts of water required, as ‘process’ water, during the production and trans- formation of goods (for example, required for the photosynthesis);
b. the amounts of water ‘embodied’ in the inputs;
c. the amount of water vapor released to air during the production and the use of com- modities;
d. the amount of water used for drinking by animals and humans, as needed in the process of food metabolism (ibid., p. 5).
In contrast, the German approach is a comprehensive one. It is oriented to a complete picture of all material (mass) flows through the economic system, but in such a way that
active and passive water are separated. In an actual and revised version of the German PIOT the primary input component comprises two corresponding water categories.
Besides, a complementary own water account was presented from the beginning of phys- ical input–output accounting.
Considering the different positions, the general problem arises, how to draw appropri- ate analytical borderlines of production processes and corresponding statistical units. In a sense, one can speak of a revival of an old debate in input–output theory concerning functional or institutional concepts of data representation.
From the point of view that ‘every production system of any type whatsoever is a system of elementary processes’ and that ‘the concept of elementary process is well defined in every system of production’ (Georgescu-Roegen 1971, p. 235), two different per- spectives are possible; on the one hand, the perspective oriented to a selected elementary (say physical and chemical) process out of the set of elementary processes that constitute the overall production process of a firm or establishment, and the perspective oriented to the overall set of elementary processes of a firm or establishment, on the other hand.
Although both perspectives are related to a functional perspective, the latter perspec- tive includes some organizational and institutional elements as is the case when an estab- lishment is chosen as a basic statistical unit. This perspective, leaving aside practical statistical aspects and recording principles, has a proper justification insofar as, for example, all water is a complementary and therefore essential input, with the consequence that the transformation process cannot take place without it. This is independent of whether passive water, say cooling water, undergoes any transformation or not. In this context, one should remember that cooling water belongs to the material input flows needed for maintaining the funds intact.
Similar problems regarding conventions, albeit with different solutions, apply to air (excluding the air mass that ‘accompanies’ the flow of used oxygen, nitrogen and carbon dioxide: see Nebbia 1999, p. 6), overburden, crude metal ores and biomass in agriculture.
(See, for example, Ayres and Ayres 1998.)
NOTES
1. A first attempt to establish a physical input–output table was made for Austria (Katterl and Kratena 1990) using input–output data for 1983. This pioneering study presented only incomplete results, especially with respect to primary inputs and final products.
2. A fund is defined as an agent in the sense of a natural or artificial system (worker, produced capital good, land) which is used and not consumed, as compared with a stock of goods which is accumulated and de- accumulated by flows. A flow is defined as a stock spread over time. A fund element enters and leaves the production process with its functional unit intact. A fund is a ‘stock of services’. (For the production theo- retical foundation of a ‘flow-fund model’, see Georgescu-Roegen 1971, ch. IX.)
3. As a complementary indicator the diagonal elements of the so-called ‘Leontief inverse’ (IA)1can be used; insofar as the diagonal elements exceed unity the existence of circuits is indicated. In the German case all diagonal elements are very close to unity and therefore circuits are absent (for methodological explana- tions, see Strassert 2001b). Generally, it should be mentioned that the results also depend on the level of aggregation. In an early version of the German PIOT with nine activities there was also a comparable high degree of linearity, nevertheless some circuits could be identified (see Strassert 2000a, p. 325).
4. Therefore, the authors continue: ‘This means that water and carbon dioxide consumed in photosynthesis, together with water vapor and carbon dioxide produced by respiration (as well as combustion) must both be included. The same is true of oxygen consumed by respiration and combination and generated by pho- tosynthesis’ (Ayres et al.2000).