Dematerialization can be the result of technological and structural changes in the use of materials (Malenbaum 1978). Technological changes imply increases in the efficiency of ti ti
Figure 18.4 Developments in the throughput index (TI)
material use through, for example, improved processes or product design. Structural changes can be defined as those changes in the composition of economic activities that have an impact on resource use. Three types of structural changes are normally mixed in the literature. They refer to (a) a change in the structure of inputs, that is, a shift in the relative shares of capital, labor and various types of natural resources in production processes; (b) a change in the structure of production, that is, a shift in the relative shares of various sectors that make up the economy; and (c) a change in the structure of consumption, that is, a shift in the composition of consumption due to changes in life styles. Both structural and technological changes can be influenced by a mix of variables, such as resource prices, governmental policies, consumer preferences and so on.
The phenomenon of dematerialization has often been explained by reference to structural changes. The intuitively appealing notion is that citizens in developing countries first show an appetite for material welfare (cars, infrastructure, consumer durables) which increases total material consumption, and that only at certain high income levels do services (banking, insurance, education, entertainment) become more important. Structural changes thus provide a logical explanation for an inverted U-shaped pattern of resource use. However, they do not explain an N-shaped pattern. The idea that consumers, in the course of economic development, start to prefer material consumption goods again, after a period in which they preferred more services, is untenable from a theoretical perspective as well as from an intuitive point of view.
However, empirical support for structural changes has been meager and unconvincing. There exists some empirical work decomposing the change in energy intensities into structural and technological factors. Howarth et al. (1991) and Binder (1993) have decomposed the change in energy intensities for a range of OECD economies. They find little support for structural changes as an important determinant of the recorded decreases in the energy intensities between 1973 and the early 1990s. The decreases in energy intensities are much better explained by referring to technological improvements in processes and product innovations. The absence of structural changes in materials demand can also be explained by reference to rebound effects. As early as 1864 the economist Jevons remarked that the savings in coal for steam engines due to technological progress are not effectuated owing to the growing demand for transport (Ko et al. 1998). Also Herman et al. (1989) have suggested that dematerialization in production may not be realized because of growing resource use in consumption. If consumers benefit financially from savings in material use in the production stage, they may spend their additional income on new consumer goods, so that the total effect can be negative. This effect was also recently demonstrated empirically by Vringer and Blok (2000).
This evidence points to the notion that patterns of resource consumption hinge critically on the development of technology over time. In economics, two conflicting views on the development of technology exist. In neoclassical economic theory, technological change follows a process of Darwinian natural selection at the margin. Whereas technological change was first assumed to be 'autonomous' and 'exogenous' to the neoclassical model, the theory of endogenous growth has more recently incorporated technological change by explicitly investigating the role of human knowledge in generating R&D and welfare. Romer (1990), for example, argues that economic growth can be enhanced by investing in 'human capital' that results in innovations and technological change. Whether innovations are rejected or accepted depends on the chances of the firm to compete more successfully in the market. Technological change is thus endogenized by making it dependent on a cost-benefit analysis concerning investments. The yields of those investments gradually improve over time because of the accumulation of knowledge. A logical implication is that the economy gradually becomes less material-intensive and more knowledge-intensive.
Alternatively, it has been suggested that the process of technological change does not follow a smooth process along a path of equilibrium, but is characterized by disequilibrium and an evolutionary path of learning and selection (Dosi and Orsenigo 1988). Innovations over time may typically come in certain clusters as the result of a process of 'creative destruction' (Schumpeter 1934). This view is supported by the accumulating evidence of biological evolution as a 'punctuated equilibrium'. Fossil evidence suggests that species remain virtually unchanged for quite a long time, but unexpected quantum leaps result in sudden appearances and extinctions. Evolution occurs not so much on an individual level but more on a macro level, and species evolve together in their environment. Gowdy (1994) applied these findings to economics and proposed that the economic system may be relatively stable and in equilibrium during a certain period of time, followed by drastic shifts (or shocks) in technological paradigms and institutional and organizational structures.
These two approaches imply a different pattern of throughput over time. The first approach seems to be compatible with the inverted-U curve in which marginal changes allow for gradually falling intensities over time. The second approach, however, is capable of explaining the N-shaped pattern in which the (temporary) dematerialization phase may be the result of a drastic shift (or shock) in technological and institutional structures. In this view the equilibrium stage may be represented by a positive linkage between throughput and income. As the result of a shock, the relationship is reversed and a period of de-materialization follows. After the effects of the shock have evaporated, the relationship returns to its equilibrium stage and rising incomes result again in rising throughput.
These two conflicting views can be investigated empirically by mapping the intensity of use in a phase diagram showing the dynamics over time. Ormerod (1994) did this for employment and found patterns of relative tranquillity in employment/GDP ratios, followed by periods of high volatility. He emphasized that various attractor points can be found in the data; the level of unemployment/GDP hovers at a certain level for a long period of time, then suddenly starts to drift until a new equilibrium level is reached. This led him to conclude that the neoclassical assumption of gradual change is not supported by the data.
Evidence for the existence of attractor points in the relationship between certain types of mass/energy throughput and income can be found in Figures 18.5 to 18.8. These figures give the development of the consumption of energy and steel per unit of GDP in the UK and the Netherlands for data in the period 1960-97 (data sources: IEA 1995; Eurostat various). The intensity of use is here plotted against two dimensions: the value in the current year and the value in the previous year. All figures show evidence of attractor points in the data. Figure 18.5 shows that, in the UK, apparent steel consumption per unit
Was this article helpful?