Consumption and the Environment

Mick Common

Centre for Resource and Environmental Studies, The Australian National University

1. Introduction

It is implicit in much of the discussion of environmental protection in the industrial economies, such as Australia, that it is a matter of changing production processes, rather than changing consumption patterns and levels. The Australian government's commitment to ecologically sustainable development does not, for example, appear to entail questioning current Australian lifestyles. Those who argue that what is required is a fundamental change in individuals' consumption habits are generally regarded as radical. For example, in the context of transportation, the mainstream arguments concern private versus public modes, leaded versus unleaded petrol, road versus rail, etc. etc. The proposition that what is required is less passenger or freight kilometres per annum is not widely canvassed, beyond the radical fringe.

We can distinguish two varieties of the view that consumption, rather than just methods of production, is, in rich nations, an environmental problem. The first identifies the pattern of consumption as the problem. This is essentially the view taken in the 'Brundtland Report' (World Commission on Environment and Development 1987). It does not call for the cessation of growth in the industrial nations. On the contrary, it explicitly argues that its continuance is necessary for the alleviation of poverty in the developing nations. However, sustainable development is taken to require changed patterns of growth in rich country consumption. On the second view, it is the level of consumption that is the problem. According to Beyond the Limits (Meadows et al 1992), for example, the level of material consumption in the industrial world needs to be substantially reduced if sustainability is to be achieved.

This distinction is useful in terms of the immediate practical implications of the positions taken. But, in some respects it is misleading and unhelpful. A changing pattern of consumption would have implications for the level of material consumption. In the transport context, for example, with passenger kilometres per annum constant, a large move from private to public modes in urban areas would reduce energy consumption, and hence a number of other environmental impacts. The level of aggregate consumption is necessarily measured in value, rather than physical, terms. Hence, it could be growing while some environmental impacts are constant or falling, due to a changing composition of the aggregate, and/or changing methods of production. However, many would argue that if the environmental impacts arising from consumption in rich nations are to be reduced so as to be consistent with the requirements of sustainable development/sustainability, then the pattern changes required are such as to mean major lifestyle changes, even if this goes with a rising value total.

In considering consumption and the environment, from the perspective of a rich industrial nation, there are three broad questions arising:

1. Given a commitment to sustainable development/sustainability, is there a need for changed lifestyles?
2. If current lifestyles are considered inconsistent with sustainable development/sustainability, does this mean that active policies are required, or can the matter be left to market forces, driven by changing consumer tastes?
3. If active policies are considered necessary, what sorts are desirable and effective?

2. Do Lifestyles Need to Change?

The answer to this question is not purely a technical matter, but depends also on the value system, or ethical stance, adopted. If we do not care at all about future generations, then the fact that current lifestyles definitely imply future impoverishment does not lead to the conclusion that current lifestyles must change. Given that we do care about future generations, as a commitment to sustainable development indicates, the answer to the question turns upon how we conceive their interests and our assessments of how current lifestyles affect those interests. This in turn depends upon assessments about substitution possibilities in production, and about the functioning of the natural environment in terms of its capabilities to provide inputs to production and to handle the wastes arising in production and consumption. These are difficult and complex questions, about which we know rather little. Some take the view that this uncertainty is itself sufficient to make the argument for lifestyle change.

Ehrlich et al (1973) introduced a very simple identity

I = P.C.T

where:

I is impact as total emissions

P is population size

C is consumption per capita

T is technology as emissions per unit consumption

Their point was that population size, affluence (the identity is sometimes written I = PAT) and technology combine multiplicatively to determine total environmental impact. The point was made to rebut Commoner's argument that technological change had been the dominant source of environmental damage in the USA in the period between WWII and 1970 (Commoner 1971). Ehrlich et al (p214) explicitly considered the example of lead emissions from automobile exhausts, where I increased by 414% over the period 1946 to 1968. The increases in the right hand side arguments were: P 42%; C, as vehicle miles per capita, 100%; T, emissions per vehicle mile, 8 1%.

This identity and example can be used to illustrate the relative roles of consumption and production. Suppose that we allow population to increase by 42% and consumption by 100%, then what is the reduction in T required to hold E constant? For P =1.42 and C = 2.00, I = 1 requires T = 0.35. To accommodate the historical population and consumption increases to constant total emissions would have required changes in the method of production of vehicle miles such that emissions per mile fell by 65%, rather than increasing by 81%. Holding population constant reduces the required emissions per mile reduction to 50%. While the identity is simple, it makes an important general point, and can in any particular case be used to figure what is required of changed production methods if increasing consumption is to be accommodated to constant, or reduced, total emissions.

3. Is Policy Action Required ?

As individuals and nations become richer, consume more, so their consumption patterns change in predictable ways. Proportionately less is spent on the material necessities of life, proportionately more on the outputs of the service sector. Given that the production of services is less energy and materials intensive, it is argued, this means that as nations get richer, so the environmental impact per $ of consumption falls. Given this, it is argued, lifestyles move in the required direction without the need for deliberate policy action.

A strong version of this argument has it that they also move at a rate such that environmental damage in total falls as economic growth proceeds. There is an argument (see, for example, World Bank 1992) that far from being a threat to the environment, economic growth is the way to save it. The basis for this argument is the 'environmental Kuznets curve' hypothesis, according to which environmental impact per capita plotted against income per capita first rises then falls with increasing income. This hypothesis has been tested by a number of researchers for a variety of impacts using a variety of data sources (notably Panayatou 1993, Selden and Song 1992, Shafik and Bandyopadhyay 1992). The results are mixed. However, even where they appear to support the hypothesis, there are a number of reasons for doubting that the implication claimed - that the world can, generally, grow out of its environmental problems - does follow (see Stern Common and Barbier 1994 for a review). The most basic consideration is that unless emissions per $ consumption go to zero at some level of consumption, in the long run emissions must grow with consumption. In the Ehrlich et al identity, an increase in C must, for constant population as P =l, imply an increase in I unless T is zero.

The idea that economic growth reduces environmental impacts per $ of total consumption relies heavily, but not entirely, on the shift out of material consumption into the consumption of services. It assumes that the production of services generates less impacts of all kinds than the production of material consumption goods. This is true, but to a smaller extent than is generally believed. The implications of increasing the consumption of service sector outputs are frequently considered in terms only of the direct inputs to, and emissions from, that sector. This is misleading, as the delivery of services uses resources and generates wastes indirectly as well as directly. It is not just a question, for example, of the direct energy inputs to broadcasting and receiving TV programmes, but also of those used in manufacturing the equipment necessary at both ends, and of those used in making the programmes shown, and the equipment used in making the programmes. For carbon dioxide emissions in Australia, for example, Common and Salma (1992) show that when indirect as well as direct responsibility is accounted for, some of the service sectors have similar emissions per $ output to some manufacturing sectors. The general point is that the delivery of the kinds of services required for current lifestyles requires a substantial material infrastructure. This limits the extent to which shifting the pattern of consumption, toward more service outputs, can reduce environmental impacts.

Actually, the explicit arguments for the environmental Kuznets curve hypothesis do not rely solely on the above income elasticities of demand type considerations. Reference is also made to production methods and changes thereto. Bernstan (1991), for example, notes that competitive pressures for cost minimisation generate incentives for reducing the use of material inputs, which, via the materials balance principle, have implications for waste discharges. It is also argued that as incomes rise, so there is an increase in 'green consumerism' (see for example Cairncross 1992), with individuals being concerned, and willing to pay more if necessary, to consume products and varieties understood to have been produced by relatively environmentally benign processes. The environmental leverage of green consumerism depends not only on the extent to which individuals are so motivated, but also on the extent to which they have the requisite information to act so as to reduce the impact of their consumption. It is sometimes claimed that much of the advertising claiming green credentials for products is misleading.

4. What Sorts of Policies?

Finally in support of the hypothesis, it is argued, in World Bank (1992) for example, that environmental protection policies are themselves a luxury good. Individual preferences are such, it is argued, that as incomes rise so does the demand for a cleaner environment, which is expressed through the political process. This 'policy demand' appears, for the population in general as opposed to activists, to relate to production methods rather than consumption patterns.

Discussion of policy instruments for environmental protection is usually mainly concerned with production rather than consumption. Those who call for lifestyle modification typically have rather little to say about policy questions, beyond generalised calls for the abandonment of the growth objective: Meadows et al (1972) and Meadows et al (1992) are examples. It is useful here to distinguish three classes of policy instrument:

1. Command and control, which involves legislating to control behaviour either in terms of allowable environmental extractions or insertions, or in terms of allowable uses of processes and equipment.
2. Moral suasion, which involves altering the social climate within which production and consumption occur, as with the provision of information on the environmental characteristics of products and processes, with publicity campaigns, as with setting ill-defined but (hopefully) striking general orientations for behaviour and policy (such as sustainable development), as with placing ill-defined injunctions on private firms and public agencies (sustainable forest management) etc. etc.
3. Price incentives, which can take the forms of: emissions taxation; input taxation; tradeable emissions, or input, permits; subsidisation of emissions abatement.

Examples of these in the private passenger vehicle use context are:

1. Design regulations for vehicles; setting allowable emissions standards; banning fuel additives.
2. Publication of fuel consumption and emissions data; negotiations with vehicle producers intended to influence their design decisions.
3. Fuel taxation; fuel taxation at differential rates on the basis of additive content; differential sales taxation on vehicles according to engine size.

The economics literature is strongly in favour of price incentive instruments: see, for example, Common (1988) for an overview of the arguments. This preference has emerged in thinking, mainly, about production rather than consumption. Until recently there was rather little use of price incentives for environmental protection anywhere. This is now changing throughout the industrial world: see James (1993) for discussion of the situation in Australia. Economists would presume that the superiority of price incentives would carry over into the consumption context. Many of those who are concerned to protect the environment and have thought about policy questions arising appear to have a predilection against the use of price incentives.

For example, the Federal government proposed, in its 1993 budget, to attack the lead additive problem by introducing a substantial differential into the tax rates on leaded and non-leaded petrol. This was opposed by the 'green' political parties. The main argument used was the regressive impact involved, in so far as older cars, alleged not to be able use un-leaded petrol without conversion, are mainly owned by the less well off. Interestingly, the argument was against the differential tax itself, not that it should be accompanied by tax/welfare changes to offset the impact on the less well off. The alternatives to differential taxation canvassed were mainly in the nature of moral suasion instruments.

Most economists would argue that changing the prices that face consumers is a very effective way of transmitting information to them, and that it is an effective means of changing their behaviour. They would doubt the efficacy of moral suasion type instruments in both respects.

While seeing a role for the provision of information, they would have reservations about publicity campaigns which sought to change peoples' preferences, which are taken as sacrosanct. It would be somewhat inconsistent to urge that policy be directed to serving the goal of consumer sovereignty, and to recommend policies intended to change the sovereign's desires. Most economists would oppose command and control type instruments on the grounds that they would be an un-necessary infringement of consumer sovereignty's range of operation, and would anyway involve avoidable costs. It should be noted, however, that some economists have argued that if sustainability is the objective, then the range over consumer sovereignty can be the criterion for social choice may have to be curtailed: see, for examples, Common and Perrings (1992), Daly and Cobb (1989).

Back...