11

even a “small” amount of climate change (if measured solely as an increase in average temperature) can trigger considerably increased probability of extreme events, particularly if the variance of the weather increases at the same time.

B. Estimates of Direct Damages

In addition to protecting against the risks of climate change, policies to reduce greenhouse-gas (GHG) emissions can also reduce some predictable damages from increasing global temperatures. Calculations of probable damages tend to take as their starting point the most firmly established consequences of global warming. such as rising sea levels, temperature-related changes in the demand for heating and air conditioning, health effects of increased temperature extremes, estimated changes in agricultural productivity, and water quality and availability. Current estimates of these potential damages are on the order of 1% to 2.5% of GDP for a doubling of atmospheric concentrations of CO2. Specific components of damage included in the calculations are listed by the IPCC (1996c) and its underlying references (Cline 1992; Fankhauser 1995; Nordhaus 1991; Titus 1992; Tol 1995). The different scholars who have made these estimates do not cover the same items in their lists of damages (although there is considerable overlap).

It is important to note what is not included in these calculations. The 1% to 2.5% of GDP range does not include damages that would result from the global warming caused by a greater than doubling of atmospheric CO2 concentrations. Yet unless action is taken, atmospheric concentrations of CO2 are projected to double by around the middle of the next century, and will continue to increase after that. Many of the damage functions underlying the estimates are non-linear, meaning that the damages could increase more than proportionally as global average temperatures rise beyond the CO2 doubling level.'

6. These are the percentages of GDP for a typical industnalized country like the United States. Damage esúmates as a fraction of GDP can be considerably higher for developing countries.

7. Atmospheric modeling calculations of the temperature increases associated with increased greenhouse gases are often benchmarked at a doubling of CO2 concentrations, but the climate change process will not come to a halt when this level is reached. Hence, estimating damages based on CO2 doubling underestimates the actual damages.

12

These damage estimates do not attempt to quantify the huge value of ecosystem services and natural capital (as calculated by Costanza et al. 1997, for example).

C. Catastrophic Possibilities

The damage estimates reported above also do not assign a value to the risk of unpleas ant surprises or catastrophic changes. By definition, the full dimensions of any such surprise cannot be known ahead of time, but three main types of potential climate catastrophe have been identified by the IPCC. All are associated with non-linear responses to increases in atmospheric concentrations of greenhouse gases: (1) a runaway greenhouse effect, (2) disintegration of the West Antarctic Ice Sheet, and (3) structural changes in ocean currents.

The "runaway greenhouse effect” refers to a situation in which the warming from human GHG emissions begins to trigger additional emissions of GHGs from biological or mineralogical systems. This would lead to a rate of climate change much more rapid than suggested by current extrapolations. Positive feedbacks might include a rapid increase in natural emissions of greenhouse gases (e.g., through methane and carbon dioxide releases from melting permafrost or methane clathrates); a shutdown of major greenhouse-gas sinks (e.g., through reduced plankton activity or the death and decay of forests), or changes in atmospheric chemistry (IPCC 1996c, p. 208). The ocean-bottom methane hydrates contain roughly as much carbon as all known land reserves of fossil fuels, and release of methane from this source may play an important role during climate change (see Dickens et al. 1997 and the references cited therein).

Recent evidence shows that the Alaska tundra region has changed from being a net sink of CO2 to a net source, raising the possibility that carbon released from the tundra could become a positive feedback to global warming (Oechel et al. 1993). Research on the response of the Antarctic ice sheets to global warming is very

13

active, with controversy between the "stabilists" and "dynanusts" over the possibility of a major melting (Sugden 1992; LeMasurier et al. 1994; Sugden et al. 1995; Horgan 1995; Schneider 1997). A number of studies suggest that changes in the pattern of ocean currents may be responsible for major climate changes measured on time scales of decades (Broecker 1995; Behl and Kennett 1996).

Even if the probability of any of these catastrophic possibilities is low, their risk value is high because the costs associated with them would be so huge. The need for prudent risk management comes to the forefront here; people are generally willing to pay to reduce the odds of unlikely but highly destructive events (residence fires, airplane crashes, nuclear power plant meltdowns). The cost of catastrophic outcomes, measured either as expected values or as peoples' willingness to pay to avoid risking them, has to be included in a complete accounting of the damages of climate change.

Finally, social and political systems would be threatened by the effects of climate change. Climate changes that reduce the habitability of low-lying coastal areas or island states could create large numbers of refugees. Forty percent of the U.S. population lives within 50 miles of the coast; it is estimated that half the world's population lives along ocean coastlines (Hanson and Lindh 1996). Similarly, if regional variations in rainfall patterns lead to desertification and localized famines, the number of climate refugees would be increased. Refugees are already a destabilizing factor in world politics, and large increases in their number would hardly contribute to peace. Because the effects of climate change would vary regionally, there is the possibility that international conflicts over water rights or other resources could be exacerbated. Climate change could destabilize domestic politics, even in democratic countries. Frightened populations might well be vulnerable to demagogic extremism in the event of an unexpected and painful climate crisis.

Of course, uncertainty runs in both directions. The actual economic value of some of the damages from climate change may be either lower or higher than their

14

expected value. Future technologies may make it easier to deal with some of the potential damages; it is not inconceivable, for example, that techniques of meteorological control might be developed that could diminish the frequency or impact of severe weather events by changing the paths of storms or influencing the locations where they drop their precipitation. These possibilities do not justify inaction, however. It also may be that future medical technology will be able to treat and cure cancer easily, but this does not mean that it would be wise policy to advise youths to begin smoking. Risk aversion implies that the distress associated with an unfavorable uncertain outcome is worse than the satisfaction of a favorable uncertain outcome, and most people are risk averse. This aversion to risk constitutes an intrinsic rationale for action now to avoid the uncertainties of climate change.

V.

THE COSTS OF REDUCING GREENHOUSE-GAS EMISSIONS

The second point in the Economists' Statement concerns the cost to the economy of reducing greenhouse-gas emissions. A complete appraisal of measures to limit GHG emissions has to include both the benefits of avoiding the risks and damages described above as well as the price that would have to be paid to achieve those benefits by reducing emissions. This is one reason Point 2 of the Economists' Statement refers to the "standard of living" rather than measured Gross Domestic Product environmental goods such as climate stability have a value as surely as the ordinary goods and services purchased in the market, although the latter are counted in GDP while the former are not. Yet the political debate, at least in the United States, has tended to focus almost exclusively on the cost side of this calculation. Special interest groups that seek to block action have made inflammatory and inaccurate statements suggesting that the cost of emissions reductions would be exorbitant. For example, the Global Climate Coalition, an industry lobby group consistently opposed to U.S. commitment to binding restrictions on carbon emissions,' recently stated that a policy to reduce CO2 emissions "would eliminate millions of American jobs, reduce America's ability to compete and force Americans into second-class lifestyles" (quotation reported in Brown 1996). Point 2 of the Economists' Statement suggests that the opposite is true: "there are policy options that would slow climate change without harming American living standards, and these measures may in fact improve U.S. productivity in the longer run." Understanding how the costs of climate change are being estimated is critical to making an informed decision about the policy options that are available.

Economists employ two distinct methodologies for estimating the cost of reducing greenhouse-gas emissions. These are broadly characterized as the "top-down" and

8. The Global Climate Coalition by no means represents the point of view of all U.S. industry. The more moderate International Climate Change Partnership (ICCP) stands for constructive engagement by industry in the formulation of the policy response to climate change, and includes as members such leading firms as 3M Company, AlliedSignal, AT&T, Boeing Company. Chevron, Dow, Dupont, Eastman Kodak, Enron, and General Electric (Inside EPA 1996; International Climate Change Partnership 1997). Other industry groups involved positively in the climate change debate include the World Business Council for Sustainable Development and the Business Council for Sustainable