AGCIEOC96/MunasingheSSSII

AGCI Session II: Characterizing and Communicating Scientific Uncertainty

Session Chairs: Dr. Richard H. Moss and Dr. Stephen H. Schneider

July 31 to August 8, 1996

Estimating Damages and Costs of Climate Change

Mohan Munasinghe

University of Colombo

Colombo, Sri Lanka

Compared to Working Group I's effort on the science of climate change, the development of Working Group III's material on the economic and social dimensions of climate change is relatively new. The 1995 IPCC WG III assessment should thus be seen as a "first cut" at the question of responses to climate change and as such, does not have the depth of material nor sufficient time to do the probing analysis of confidence intervals that we see in the Working Group I report.

Predictions about climate change, its impacts, and the costs of mitigation are important for the policy making dimension. We are not seeking numbers for their own sake. Questions about climate change reside within broader questions about sustainable development. The objective of human development is sustainability, and the pursuit of greater precision in climate prediction can help with progress toward this goal.

While climate change has an important bearing on sustainable development, it is crucial to recognize that, especially for the developing countries, there are a number of other priorities. Hunger and malnutrition, poverty, and pressing local environmental issues present far more immediate needs. For example, particulates from burning fuelwood in traditional stoves expose large numbers of women and children to the equivalent of smoking 20 packs of cigarettes a day. And in middle income countries, sulfur dioxide pollution, especially in capital cities, is a much higher priority than climate change.

In terms of priority setting and getting the attention of decision makers in low and middle income countries, it is important to have information emerge from the IPCC process to show how climate change is related to these other priorities. For example, the most attractive climate change policies would be so-called "win-win" options, such as energy conservation, which provide significant in-country benefits as well as reductions in greenhouse gas emissions. It is also important to consider equity issues, like the huge disparity in per capita carbon dioxide emissions between the high and low income countries, when decisions are made concerning future international burden-sharing to mitigate climate change.

While climate change has an important bearing on sustainable development, it is crucial to recognize that, especially for the developing countries, there are a number of other priorities.

Some of the less developed countries are beginning to pay attention to the issue of climate change however, and Munasinghe believes things have progressed considerably in this regard. His own country, Sri Lanka, has at least three reasons to be concerned about climate change, he explains. In the short term, the UN Framework Convention on Climate Change (FCCC) will involve joint implementation projects that could earn the country foreign exchange currency. In the middle term, as the FCCC comes increasingly into force, it could affect decisions regarding the building of power plants in Sri Lanka and have impacts on other aspects of sustainable development. In the long term, future generations will be affected worldwide. How we pose the problem and solutions is important. To get attention, we must focus on certain near-term issues, embedding them in a long-term context.

Regarding key uncertainties in Working Group III's assessment, Munasinghe says that the range of confidence in the predictions is very loose. Mainly, the results consist of "best guess" estimates based on the judgments of the so-called experts who wrote the chapters. No error bands can be assigned and there are no probabilities attached to particular outcomes.

The chain of causality from emissions to impacts to responses is summarized in Figure 2.13. Hopefully, through the FCCC, the allocation of rights and responsibilities will close the loop in a way that reduces net emissions and also controls impacts and damages.

It is important to consider equity issues, like the huge disparity in per capita carbon dioxide emissions between the high and low income countries, when decisions are made concerning future international burden-sharing to mitigate climate change.

The decision framework roughly breaks down into three major components, which are not mutually exclusive:

1. global optimization is the most important component, which looks at costs and benefits and averages and aggregates them to find some optimal level for reduction of GHG concentrations, adaptation responses, and so on;

2. collective decision making, which covers international bargaining, equity issues, and other such matters that cannot be captured in global optimization models;

3. mechanisms and procedures, which deals with process issues, including creating or adapting international bodies, developing monitoring mechanisms, etc.

The most attractive climate change policies would be so-called "win-win" options, such as energy conservation, which provide significant in-country benefits as well as reductions in greenhouse gas emissions.

There are many complications in the analysis of global climate change. First, the tools for such an analysis have never been used on such large spatial and temporal scales. Similarly, the complexity of this issue is far greater than any other tried previously. The issue of "irreversibility" causes another complication we may lose future options which have value. What would we be willing to pay to have the flexibility of exercising certain options in the future? Non-linearity, or the possibility of catastrophic outcomes, adds another complication to the analysis. And, directly on the subject of this meeting, uncertainty is a key complication. We have, in this analysis, socioeconomic uncertainties that overlay all the scientific uncertainties, compounding the problem. Finally, we have equity and social issues, which have a strong influence on the negotiating process and the final outcome, but are very difficult to factor into the analysis especially in the decision making stage.

In the cost/benefit approach used to determine the globally optimal level of emissions, cumulative emission reduction is plotted against the cost of removing each ton of carbon from the atmosphere. Analogously, the marginal benefit curve is drawn based on the damage one avoids by removing one ton of carbon from the atmosphere. Theory says that we should carry out emissions reduction up to the point where the cost of removing the last ton of carbon is equal to the damages avoided (where the two lines cross). The problem is that we do not know the shape of these curves.

We have, in this analysis, socioeconomic uncertainties that overlay all the scientific uncertainties, compounding the problem.

Suppose the damage curve was poorly defined, but we still had some idea of the costs of mitigation. Then, instead of the optimal reduction level approach mentioned above, we could use the "affordable standard" (which implies that we reduce emissions up to the point that we can afford to, say 1 percent of GDP per year). A more stringent criterion may be based on the "absolute standard" approach, which indicates that there is some benchmark or point beyond which any transgressions will result in very steeply rising costs and high risk of some catastrophic outcome.

Damages involve valuation of ecological impacts such as biodiversity loss which are very difficult to estimate and have a much higher order of uncertainty. Also, damages involve potential social impacts which may be so severe in some places that they damage the whole social fabric causing mass migrations, civil strife and conflicts that are impossible to quantify. For these reasons, Munasinghe feels that damage estimates are inherently less precise than mitigation costs estimates (which are technology dependent).

Table 2.14 summarizes the monetary damage estimates for a doubling of CO2 on an annual basis. Ranges are upper and lower bounds based on the existing literature; no error bars are assigned here. Assumptions are for 560 ppmv CO2 equivalent and a 2.5°C climate sensitivity.

Munasinghe and colleagues feel that damages involve many more uncertainties and are far more difficult to estimate than costs, and as such, they assigned wider error bars to their damage estimates.

The damage costs range from $5 to $125 per ton of carbon emitted. Different values of discount rate (based on the notion that costs in the future should carry less weight than costs borne today), which is used in various calculations, cause damage costs to vary substantially. If a uniformly high discount rate is used, say 5 percent, the range of damage costs narrows to $5 to $12 per ton of carbon. However, environmen -talists and others argue that with a problem like climate change, we should use a very low discount rate implying that future damages should weigh heavily in the calculus today. What discount rate to use is the source of great controversy. While the choice of a discount rate is a source of dispute and not another cause of uncertainty, it does influence estimated damage costs significantly.

Regarding mitigation costs, some experts believe that CO 2 could be reduced to 1990 levels at negligible costs, using "bottom up" models which deal with individual behavior. Conversely, the top down, macroeconomic models project the costs of mitigation to be much higher, up to 1-2 percent of GDP. Nevertheless these top-down models give answers for different countries that vary quite widely. To hazard a guess from these results is a perilous task indeed. However, for decisions that must be made today, we should strive to give decision makers information on what is currently known, and at the same time seek to improve the state of our knowledge for future assessments. Climate change decision making is not a single event, but rather a process based on a heuristic approach which makes use of continuing improvements in human knowledge.

Some experts believe that CO2 could be reduced to 1990 levels at negligible costs, using "bottom up" models which deal with individual behavior. Conversely, the top down, macroeconomic models project the costs of mitigation to be much higher, up to 1-2 percent of GDP.

Forward to Next Section//Back to Table of Contents//AGCI Homepage//Comments: agcimail@agci.org