Discounting and the Environment Should Current Impacts be Weighted Differently than Impacts Harming Future Generations?

Background. In Life-Cycle Assessment (LCA), decision makers are often faced with tradeoffs between current and future impacts. One typical example is waste incineration, where immediate emissions to the air from the incineration process have to be weighted against future emissions of slag landfills. Long-term impacts are either completely taken into account or they are entirely disregarded in case of a temporal cut-off. Temporal cutoffs are a special case of discounting. Objective. In this paper, discounting is defined as valuing damages differently at different points of time using a positive or negative discount rate. Apart from temporal cut-offs, discounting has rarely been applied in LCA so far. It is the goal of this paper to discuss the concept of discounting and its applicability in the context of LCA. Methods. For this purpose, we first review the arguments for discounting and its principles in economic sciences. Discounting in economics can be motivated by pure time preference, productivity of capital, diminishing marginal utility of consumption, and uncertainties. The nominal discount rate additionally includes changes in the price level. These arguments and their justification are discussed in the context of environmental impacts harming future generations. Results and Discussion. It is concluded that discounting across generations because of pure time preference contradicts fundamental ethical values and should therefore not be applied in LCA. However, it has to be acknowledged that in practice decision makers often use positive discount rates because of pure time preference – either because they might profit from imposing environmental damage on others instead of themselves or because people in the far future are not of immediate concern to them. Discounting because of the productivity of capital assumes a relationship between monetary values and environmental impact. If such a relationship is accepted, discounting could be applied. However, future generations should be compensated for the environmental damage. It is likely that they would demand a higher compensation if the real per capita income increases. As both the compensation and the discount rate are related to economic growth, the overall discount rate might be close to zero. It is shown that the overall discount rate might even be negative considering that the required compensation could increase (even to infinite) if natural assets remain scarce, whereas the utility of consumption diminishes with increasing Introduction It has long been recognized that certain environmental decisions involve tradeoffs between present and future impacts. Such tradeoffs raise issues of (intergenerational) fairness and equity that are ethical in nature. Life-Cycle Assessment (LCA) involves many of such temporal issues [1]. For instance, construction materials are often 'stored' in buildings for many decades before they are recycled or disposed of [2]. It is uncertain, which disposal technologies will be used for these materials in the future. Information about the composition of materials might be lost in the course of the years, which makes an adequate treatment more difficult. The use of some of these materials or substances with a high impact or risk potential might have even been forbidden or limited meanwhile (e.g. CFC), but they remain to be released to the environment for long times. Resources are bound in the building that cannot be used for other purposes. While the initial LCA Methodology with Case Study Discounting and the Environment Int J LCA 8 (1) 2003 9 use of resources is usually considered in LCA, the time period of use is neglected [2]. Moreover, when materials are recycled, impacts are generally not allocated to the secondary product in LCA. Finally, some of the emissions, e.g. from land-filling construction residues, occur in the far future and it is unclear how they should be assessed. In general, LCA makes no explicit differentiation between emissions (and, ultimately, impacts and damages) at different points in time. For instance, whether an emission contributes to ozone depletion today or in 200 years is treated equally in LCA. Nevertheless, there are some forms of implicit discounting that are common practice, e.g. temporal system boundaries [3]. Temporal cut-offs are a special case of discounting with a discount rate of zero for the time horizon considered and of infinity thereafter. Such temporal system boundaries are often proposed for landfill emissions [4,5]. Similarly, the distinction between short-term emissions (≤100 years) and long-term emissions (>100 years) [6,7] implies a different valuation for the two periods. In the impact assessment phase, the choice of several models is related to time. For instance, different time horizons can be chosen for the global warming effect, typically 20, 100, and 500 years. Depending on this choice, the global warming potential of a pollutant may differ considerably. Huijbregts et al. [8] found that metal toxicity potentials differed up to 6.5 orders of magnitude depending on the time horizon chosen in the fate model. These large differences are due to long residence times and slow removal pathways from the soil and marine water/sediment compartments. To reduce the importance of long-term emissions, several methods introduced fictitious 'removal rates' or 'degradation rates' for metals in the fate modeling, (another example of implicit discounting) [9,10]. Despite these examples, and in contrast to other projects such as the Impact Pathway Method of the ExternE Project1 [12], explicit discounting has rarely been applied in LCA (exceptions are [13,14]). Since LCA is a valuebased decision support tool, it needs to address time-preferences if they are relevant in the context of future environmental damages. There are some applications such as waste incineration, where tradeoffs between impacts in the present and in the future have to be made. For instance, new thermal waste treatment technologies prevent long-term emissions at the cost of a higher energy use with more immediate impacts. These issues demand a thorough discussion of whether impacts at different points in time should be weighted alike and, therefore, whether discounting (including temporal system boundaries) should be applied in LCA. In this paper, we briefly introduce the economical background of discounting (Section 1). Emphasis is placed on motivating the use of discounting in LCA (Section 2) and on illustrating the consequences of discounting in a case study (Section 3). It is neither the intention of this paper to discuss all ethical standpoints involved (see [3] for an overview) nor to provide a universal answer of whether or not discounting should be applied in the context of LCA. An anthropocentric point of view is adopted throughout the whole paper (as is always the case in economic discounting). Furthermore, it is assumed that weighting between different environmental impacts, as often performed in quantitative LCA, is not perceived as 'unethical'. Otherwise, weighting environmental damages at different points of time is certainly also not accepted so that the following discussion would be needless. 1 The Concept of Discounting: Economical Background In the economic sciences, future costs and benefits are usually discounted to a present value in order to make them comparable to current costs and benefits (cost / benefit analysis) [15]. There are several reasons for valuing one monetary unit of benefit or cost differently at different points of time. For instance, pure time preferences (impatience), the productivity of capital (related to economic growth / decline), and uncertainty / risk perception are all factors that change valuations with time. The way the valuation changes in time is generally known as the discount rate. This rate depends on the factors just mentioned. Thus, in economics, the Net Present Value (NPV) of an investment is calculated as a function of benefits, costs, and the discount rate (Equation 1):