Capital stock, capital stock turnover and new investments

Mitigation costs depend on the lifespan of existing plants and equipment. The lifespan for energy producing and using capital stock (for example, power plants, housing and transport) is not fixed. It is influenced by factors such as maintenance costs and reliability, which tend to change over time. Nevertheless, energy-related capital stock is typically long-lived and premature retirement is apt to be costly. To avoid premature retirement, mitigation efforts can be spread more evenly over time and space. To reduce the cost of any stabilization target, SAR WGIII stresses the need to focus on new investments and replacements at the end of the economic life of plant and equip ment (i.e., at the point of capital stock turnover).

The focus on new investment does not imply "doing nothing". Acting too slowly- not even undertaking low cost measures — may increase the costs of a stabilization path by requiring more rapid action later on. This may include the need to retire, prematurely, capital stock that is constructed in the interim. For example, deferring mitigation for a couple of decades would allow global fossil fuel emissions to increase significantly (e.g., IS92a and several other scenarios). But to stabilize concentrations below 450 ppmv, emissions would have to be brought back down to 1990 levels by about 2040 and lower thereafter. This might require society to replace much of the stock constructed in the interim, and these costs need to be weighed against any economic benefits gained from the deferment.

The optimal rate at which capital stock is replaced reflects broader questions about the inertia of energy systems. For example, different investments have different time implications. Constructing new, very long-lived, carbon-intensive infrastructure may raise the costs of limiting emissions many decades from now. Discouraging investments such as inefficient buildings, or other urban infrastructure that may encourage a wide range of carbon-intensive activities, could be important now in lowering the long-run costs of stabilizing atmospheric concentrations even at higher levels. However, the issue of inertia and bow it affects different investments is not well understood.

As indicated by Figure 5, a 450 ppmv limit would require reductions in global emissions starting very soon, while higher limits would delay the need for restrictions. While emissions increases in some countries can be offset by declines within others over some period of time, emission growth must eventually be curtailed in all regions to meet the limit.

Technical progress

35

Abatement costs at any point in time rise with the quantity of emissions abated at that time. The suite of abatement technologies described in SAR WĢII can be considered as forming a "supply curve". Clearly, it is cheapest to take the least-expensive measures first and to work up the "supply curve" using more costly measures as required to meet the objective.

Technical change is likely to reduce abatement costs over time. The rate of this reduction may depend on the stabilization level and emission pathway. Stabilization levels and emission pathways that imply more immediate reductions may stimulate development of new, lower carbon technologies: "induced" technology development. This increases long-run flexibility and lowers the long-run costs of a carbon constraint, but at a nearterm price. According to this argument, rather than wait for technology development to lower future mitigation costs, early emission constraints induce the private sector to undertake appropriate research and development, including the switching of research and development investment away from exploration and development of carbon-intensive resources and technologies.

Induced (endogenous) technical change depends on the stimulation of innovation by price signals, which is likely to be greatest in well functioning markets. In the early stages of technology development, it is difficult to establish ownership of research results; therefore the private sector often is reluctant to invest in adequate research and development. The prospects of future markets is unlikely to overcome this problem entirely. This well known market failure is often used to justify government involvement in research and development, and such research and development may be very important in promoting the development of technologies early on.

Government research and development and emission constraints are not the only levers policy makers can exercise to influence the rate of technology development, diffusion and dissemination. Tax incentives and the support of "protected" markets, such as premium payments for renewable energy, may also encourage the private sector to invest in carbon-free energy and the development of associated industries. Technology diffusion and dissemination may also be inhibited by market failures and require specific policies to overcome.

In reality, a mix of all these measures - greatly increased government research and development, support for technology distribution, explicit market supports, and appropriate emission constraints probably will act together to stimulate the technology needed to lower the costs of stabilizing atmospheric CO2 concentration. The literature assessed in SAR WGIII does not give a clear indication as to the appropriate mix of policies and the implications for emission pathways.

The least expensive mitigation options are often associated with new investments. To take advantage of these opportunities, a

36

cost-effective approach would adopt low cost mitigation measures wherever new investments are made throughout the world. Mechanisms such as emissions trading or joint implementation may be used to implement this strategy in a manner that facilitates the distribution of mitigation costs among countries while promoting cost effectiveness. This approach, commonly referred to as "where" flexibility, works because the climate benefits of CO2 emission reductions do not depend on their location.

Discount rate

With regard to mitigation costs (the subject of this section), a positive discount rate lowers the present value of the costs incurred. This is because it places a lower weight on investments made in the future. Indeed, the further in the future an economic burden (here, emission reductions) lies, the lower the present value of costs. In a wider context, discounting reduces the weight placed on future environmental impacts relative to the benefits of current energy use. Its use makes serious challenges, such as rapid switching of energy systems in the future, seem easy in terms of present dollars and may affect consideration of intergenerational equity.

Carbon budget

Carbon emissions may follow different pathways to meet a certain stabilization target (as shown by Figures 5 and 6). If no major disruption of the processes that govern the uptake of CO2 by the ocean and the land biosphere occurs, then long-term total cumulative emissions for a given stabilization pathway are essentially independent of the pathway towards a stabilization target (see Figure 6 and Section 2.2). However, the allocation of emissions in time depends on the pathway. Emissions in the next decades can be notably higher for pathways that follow IS92a initially (see Figures 6 and 7). Thus, the requirements for higher cost carbon-free alternatives are reduced in the short-term and stronger emission reductions are delayed into the future.

However, there are risks associated with emission pathways that follow IS92a initially. Higher earlier emissions and implied higher concentrations and rates of concentration increase may disrupt the physical and biogeochemical processes governing the flow of carbon. This may mean that emissions must be lower than expected to meet a certain stabilization target. In addition, higher earlier emissions will lead to faster rates of climate change, which may be costly. Pathways that imply higher emissions initially may have a more rapid transition from increasing to decreasing emissions, which tends to increase mitigation costs.

next decade. Projections over a century or more must be treated with considerable caution. Nevertheless, such exercises can provide useful information. The value however, lies not in the specific numbers, but in general results that are useful for policy making.

3.2.2.1 Studies Available at the Time of the SAR WGIII

Until recently, proposals for dealing with climate change tended to focus on emissions rather than concentrations: for example, returning emissions to 1990 levels by 2000, or a 20 per cent reduction by 2005. As a result, few analyses had examined the economics of stabilization at the time of SAR WGIII. Those that bad are reviewed in Chapters 9 and 10 of SAR WGIII and are described below. (Subsequently a number of additional studies have been undertaken, but, in accordance with the guidelines for Technical Paper preparation, they are not reviewed here.)

Several authors have explored the cost-effectiveness of a particular CO2 concentration target. For example, Nordhaus (1979) and Manne and Richels (1995) identify least-cost mitigation strategies for meeting a range of alternative concentration targets. They found that the least-cost mitigation path initially involves modest reductions from the emissions baseline. Higher concentration targets allow emissions to follow the baseline for longer periods.

Richels and Edmonds (1995) and Kosobud, et al., (1994) examined alternative emission pathways for stabilizing atmospheric concentrations. Their results indicate that pathways involving modest reductions in the early years, followed by sharper reductions later on, are less expensive (in terms of mitigation costs) than those that require substantial reductions in the short-term given their assumptions concerning technical change, capital stock turnover, discount rate and the effect of the carbon budget. The timing of emission reductions is known as "when" flexibility.

Higher stabilization targets allow more flexibility in the rate of departure from the baseline. However, regardless of the rate of departure from the baseline, a stabilization pathway is not a "do nothing" or "wait and see" strategy. First, each concentration path still requires that future capital equipment be less carbon-intensive than under a scenario with no carbon limits. Given the long-lived nature of energy producing and energy using equipment, this has implications for current investment decisions. Second, new supply options typically take many years to enter the marketplace. To have sufficient quantities of low cost, low carbon substitutes in the future would require a sustained commitment to research, development and demonstration today. Third, any available no-regrets measures for reducing emissions are assumed to be adopted immediately, which may require government action.

Two aspects of the above studies arouse considerable debate: the goal, and the reliance on highly simplified models of the energy-economic system. With regard to the former, the authors stress that their focus has been on mitigation costs, with particular attention to the least-cost path for meeting a particular concentration target. They emphasize that it is also important to examine the environmental consequences of choosing one emission path over another. Different emission paths imply not only different mitigation costs, but also different benefits in terms of averted environmental impacts, as well as the injection of novel environmental issues, such as those that might occur if biomass fuels become more important.

The analyses are also limited by their treatment of uncertainty. Uncertainty regarding the ultimate target is likely to persist for some time. Under these conditions, policy makers must identify a prudent near-term hedging strategy that balances the risks of acting too slowly against the costs of acting too aggressively. Although several of the studies cited in SAR WGIII attempt to assess the robustness of the near-term control decision to the long-term concentration target, they do not analyse the effects of uncertainty explicitly.

Some critics also dispute the methodologies that underlie these studies. They question the extent to which the models, which by necessity simplify the energy-economic system, capture the full complexity of capital stock, its interlinkages and other sources of inertia in the system. For example, existing models do not simulate the linkages among investments. Some investments we take today, like roads, last for a very long time and create a whole network of interlocking investments (e.g., the spatial pattern of industrial facilities and housing) that may affect the costs of emission constraints for years to centuries.

37

between emissions consistent with a given stabilization target and some baseline, ignoring the ecological or marine feedbacks can increase or decrease the emissions and mitigation costs associated with a stabilization level. Given the scientific uncertainties in the carbon models, the uncertainty from oceanic and terrestrial feedbacks is likely to be ±100 GtC or more.

In practice, we do not know the appropriate stabilization level, and this makes the appropriate strategy still more complex. Stronger research and development policies, which are relatively cheap compared with the potential costs of rapid reductions in emissions, appear a good investment against a wide range of outcomes. In addition, early mitigation, particularly at the point of new investment, reduces the exposure of the economy to the possibly very high costs of discovering that we need to achieve a lower stabilization target than expected initially. Fuller implementation of no-regrets and low cost measures help, not only to reduce impacts, but also to prepare economies for stabilization.

As noted earlier, sensible greenhouse policy requires decision makers to consider the costs and other implications of climate change policy measures together with what such measures might buy in terms of reducing the undesirable consequences of global climate change. In Section 3.1, we discussed the issue of impacts and how they may be reduced by adopting a lower stabilization target. In Section 3.2, we discussed mitigation costs associated with limiting anthropogenic CO2 emissions to achieve stable atmospheric concentrations. This section discusses possible insights from integrating this and other relevant information contained in this paper.