- Email: [email protected]
Economics, ethics, and climate policy: framing the debate
GLGBAL ANJIPGTETABY
ELSEVIER
Global and Planetary Change
11( 1996) 187- 199
Economics, ethics, and climate policy: framing the debate
’
Richard B. Howarth a, Patricia A. Monahan b a Environmental Studies Program, University of California, Santa Crut, CA 95064, USA b Energy and Environment Division, Lawrence Berkeley Laboratory, Berkeley, CA 94720, USA Accepted
30
November 1995
Abstract This paper examines the economic and ethical dimensions of climate policy in light of existing knowledge of the impacts of global warming and the costs of greenhouse gas emissions abatement. We find that the criterion of economic efficiency, operationalized through cost-benefit analysis, is ill-equipped to cope with the pervasive uncertainties and issues of intergenerational fairness that characterize climate change. In contrast, the concept of sustainable development-that today’s policies should ensure that future generations enjoy life opportunities undiminished relative to the present-is a normative criterion that explicitly addresses the uncertainties and distributional aspects of global environmental change. If one interprets the sustainability criterion to imply that it is morally wrong to impose catastrophic risks on unborn generations when reducing those risks would not noticeably diminish the quality of life of existing persons, a case can be made for significant steps to reduce greenhouse gas emissions.
1. Introduction Debates over the appropriate policy response to global climate change involve fundamental disputes concerning social values and the role of science in policy analysis. Some argue that policies should equate the marginal costs and benefits of greenhouse gas emissions as defined in economic terms. Others hold that climate change involves ethical dimensions that are not easily incorporated in economic analysis -in particular, that future generations have a moral right to a stable climate that imposes binding obligations on present actions.
’ This article is a revised and much-condensed version of the report “Economics, Ethics, and Climate Policy” (Howarth and Monahan, 1992). We thank the Stockholm Environment Institute for fmancial support. 0921-8181/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved SSDI 0921-8181(95)00052-6
If we were certain that the impacts of climate change would be gradual and easily managed by social adaptation, reducing greenhouse gas emissions would hardly appear as an urgent policy priority. If, on the other hand, we were certain that unmitigated emissions would imply catastrophe for the well-being of future societies, few would argue for a “go slow” approach to policy intervention. But although scientists have reached broad agreement on a number of issues relating to climate change, a key fact is that both the dimensions of climate change and its impacts on human and ecological systems will remain highly uncertain. Coping with uncertainty is therefore a central policy challenge. This paper examines alternative frameworks for analyzing climate change as a global policy challenge with a special focus on cost-benefit analysis and principles of intergenerational fairness. We find that the criterion of economic efficiency as opera-
188
R.B. Howarth, PA. Monahan/Glohal
tionalized through cost-benefit analysis is poorly suited to the evaluation of climate policy for two fundamental reasons: (1) cost-benefit techniques are difficult to operationalize given the gross uncertainties surrounding climate change and its potential effects on future society; and (2) the approach is insensitive to the distribution of impacts between present and future generations, yet climate change may have pronounced implications for future economic conditions and environmental quality. Under the rubric of “sustainable development”, intergenerational fairness has been widely embraced as a criterion in environmental planning. We explore the ethical basis of the sustainability approach, its translation into operational criteria, and its implications for climate policy. The sustainability criterion may be taken to imply that it is morally wrong to impose catastrophic risks on members of unborn generations if reducing those risks would not significantly diminish the quality of life of existing persons. Under this interpretation, a case can be established for significant greenhouse gas emissions abatement given current knowledge of mitigation costs and the risks of climate change. Section II sketches the key facts and uncertainties of climate science as they relate to policy formulation. Section III outlines the potential impacts of climate change on human and environmental systems and the difficulties of translating those impacts into economic terms. Section IV examines the cost of policies to abate greenhouse gas emissions, while section V explores the logical foundations of costbenefit analysis and the principle of sustainable development as guides to public policy. We conclude by reviewing the principal findings that stem from our analysis.
2. Climate science: the emergent facts A working consensus has been reached regarding key aspects of climate science. Atmospheric concentrations of greenhouse gases [carbon dioxide, chlorofluorocarbons (CFCs), methane, nitrous oxide, and water vapor] have increased rapidly due to human activities such as fossil fuel combustion, refrigeration, air conditioning, deforestation, and agriculture. In the absence of steps to reduce greenhouse gas
and Planetary Change 11 (1996) 187-199
emissions, a doubling of atmospheric CO,-equivalent relative to the pre-industrial level is anticipated by 2025 (IPCC, 1991a). General circulation models (GCMS) project that a doubling of CO,-equivalent would raise the Earth’s mean temperature by 1.9-5.2”C, with 2.5”C as the best estimate (IPCC, 1991b). While such changes may seem small when compared against daily and seasonal variations, a change of only 1°C heralded the Little Ice Age between the 14th and 17th Century, inducing frequent crop failures and sporadic freezing of the Baltic Sea (Oeschger and Mintzer, 1992). A 5°C temperature increase would move the Earth to a climatic regime not experienced in over a million years. Beyond these essentials, scientific research points to a range of potential impacts that climate change might entail, from enhanced agricultural yields to the inundation of island nations. Given the prevailing state of knowledge, neither the probability nor severity of climate impacts can be gauged with confidence. There is, however, general agreement in the scientific community that the potential for catastrophic change exists. This section outlines some of the more likely impacts of climate change, as well as a number of catastrophic outcomes that might result from feedback processes that are not well understood by existing science. As temperatures warm over the course of the next century, sea level is expected to rise by up to one meter. Since half of humanity currently occupies coastal zones, a change of this magnitude would have far-reaching effects. Most at risk are poorer nations with densely populated coastal areas. Indonesia, with 15% of the world’s coastlines, is projected to lose 40% of its land surface (Schneider, 1989). A one-meter rise could inundate 15% of Bangladesh and all of the Republic of Maldives, Kiribati, the Marshall Islands, Tokelau, Tuvalu, and the Torres Strait Islands (Hulm, 1989; Lewis, 1989). Rapid climate change would alter the composition of ecosystems, with some species benefiting and others unable to migrate or adapt at the rate necessary for survival. Given a warming of O.l-1°C per decade, Woodwell (1990) finds that the capacity of natural communities to migrate is exceeded by factors of 100-1000 or more; a substantial loss of biodiversity is a distinct possibility. Impacts on hu-
R.B. Howarth. P.A. Monahan/Global
man systems will likely be felt most acutely in regions already under stress, such as those exposed to the natural hazards of coastal and riparian flooding, drought, land-slides, severe storms, or tropical cyclones. Climate change might also increase the frequency and intensity of tropical storms; diminish water supplies in areas already experiencing water shortages (Gleick, 1992); force the mass migration of human populations; and reduce agricultural yields, particularly in low-income regions unable to adapt to changing climatic c’onditions. Today’s GCMs do not fully incorporate feedback processes between atmospheric, oceanic, and biogeochemical systems. But as Schimel (1990, p. 68) points out, “Feedbacks between atmosphere and biosphere are non-linear, sensitive to initial conditions, and capable of enormous amplifications. Complex feedbacks in the Earth system can produce unexpected and potent responses... Without crying wolf, it is worthy of our concern as a society that biogeochemical and ecological feedbacks may result in more rapid environmental change than is predicted by purely physical models.” Some 10,000 years ago, a shut-down of ocean circulation may have contributed to regional temperature increases as large as 7°C and a 50% increase in rainfall over the course of a few decades (Dansgaard et al., 1989). Broecker (1987) suggests that changes in temperature and precipitation patterns may alter the circulation of the world’s oceans, inducing unpredictable but potentially catastrophic feedbacks. A second potential feedback is the accentuation of greenhouse gas emissions from the biosphere. Global warming could accelerate the anaerobic decay of organic matter, stimulating methane generation, as well as melting methane clathrates stored in Arctic sediments. Revelle (1983) estimates that clathrate liberation might amplify global warming by 0.6518°C. Although such results point to the potential for catastrophic change as a key fact of climate science, neither the probability nor consequences of such outcomes is easily quantified. Yet as Svedin and Aniansson (1987) warn, by “leaving out the external shocks, nonlinear responses, and discontinuous behavior so typical of social and natural systems, sur-
and Planetary Change 11 (1996) 187-199
189
prise-free analysis leaves us unprepared to interpret a host of not-improbable eventualities”.
3. The economic costs of climate change The application of cost-benefit analysis to the assessment of climate policy requires monetary estimates of the costs and benefits of climate change. In principle, monetizing costs and benefits would allow for the comparison of dissimilar impacts in a consistent framework to identify an efficient level of greenhouse gas emissions abatement. The rigorous application of cost-benefit analysis, however, requires information on the full range of possible outcomes and their respective probabilities. Since neither can be identified with confidence given the prevailing state of scientific knowledge, economic analysis is unable to provide unambiguous estimates of the relevant costs and benefits. Using the same base of scientific information, Nordhaus ( 19891, Nordhaus, 1990, 199 1a> and Cline (1992a) Cline, 1992b) derive rather different estimates of the damages a doubling of atmospheric CO,-equivalent would impose on the U.S. economy (Table 1). A comparison of these studies lends insight into how analysts’ worldviews and intuitions enter the valuation calculus. Agricultural impacts reflect the greatest divergence between the Nordhaus and Cline damage estimates. While elevated CO, concentrations and longer growing seasons enhance plant yields under laboratory conditions, shifts in soil moisture (Parry, 1990), increased range of pests and diseases (USEPA, 1990), and changes in the probability of extreme weather events (Meams et al., 1984) might lead to reduced yields. Nordhaus expects the costs and benefits to agriculture to balance at $10 billion each with zero net costs. This figure reflects a USEPA (1989a, , 1989b) estimate that assumes a greater carbon fertilization effect than will be realized with a doubling of CO,-equiualent. Cline tempers USEPA’s estimate to account for a lower carbon fertilization effect and includes the costs from increased incidence of severe drought (Rind et al., 19901, estimating net costs of $8.6 billion for the 1981 breakdown of the economy. Nordhaus and Cline also differ in their approach to nonmarket impacts such as species extinction and
190
R.B. Howarth, P.A. Monahan/Global
Table 1 Damage costs to the U.S. economy for a doubling of CO,-equivatent (billion 1981$ per year) Study
Cline, 1992a a
Nordhaus, 1991a
1981 National income Forestry Electricity/Energy Air conditioning Heating Sea level rise Loss of land Coastal protection Recreation Ozone impacts Health Water supply Agriculture Species loss Migration Hurricanes Urban infrastructure Total of 1981 income
2415.1 1.6
2415.1 small
5.5 -0.6
1.65 - 1.16
2.9 0.6 0.8 1.7 2.9 3.5 8.6 2.0 0.2 0.4 0.05 30.1 1.25%
1.5 3.7 small not quantified not quantified not quantified -9.7-10.6 not quantified not quantifted not quantified not quantified 6.2% 0.25%
a Cline’s initial estimates were based on 1990 prices and economic conditions. We multiplied these figures by the ratio of nominal economic activity in 1981 and 1990 to obtain the estimates reported here.
loss of human life in natural disasters. Nordhaus (1991a, p. 44) acknowledges that nonmarket goods “escape the net of the national income accounts and might affect the calculations” but notes that although “[some] people will place a high moral, aesthetic, or environmental value on preventing climate change,... I know of no serious estimates of what people are willing to pay to stop greenhouse warming”. Cline, in contrast, explicitly examines the potential costs of impacts on recreation, water supply, species loss, migration, hurricane damage, urban infrastructure, health and welfare, and forestry. He finds these costs to be cumulatively greater than the costs of climate change on agricultural output. Despite these differences, Cline and Nordhaus work with closely similar damage cost estimates (1 .O vs. 1.25% of economic activity for a doubling of CO,-equivalent) when integrating the costs and benefits of greenhouse gas emissions abatement. Although Nordhaus’ direct cost estimate is only onefifth of Cline’s figure, he judgmentally adjusts his cost coefficient from 0.25 to 1.0% to account for
und Planetmy
Chunge II (1996) 187-199
omitted factors (Nordhaus, 1992). Both Cline’s and Nordhaus’ cost estimates are plausible based on existing science, yet each relies strongly on subjective judgements that are, at present, not amenable to empirical test. Much larger impacts might arise if climate change turns out to be worse than expected; smaller impacts might also occur. Since the key uncertainties are not quantified, such estimates provide an insufficient basis for the rigorous application of cost-benefit techniques. A final issue concerns the functional relationship between greenhouse gas concentrations and related economic damages. The literature focuses mainly on the expected impacts of a doubling of atmospheric CO,-equivalent, yet the potential impacts of much larger concentrations are of key policy relevance. Cline (1992b, p. 376) employs a geometric damage function under the premise that the “greenhouse effect poses major risks, especially over the very long term of two to three centuries, by which time temperatures could rise by as much as 10” to 18°C”. He specifies climate damages per unit of economic activity (d) as d= d,, x ( AT/ATo)1.3
(1)
where AT is the temperature increase caused by climate change while d, and AT, are the damage level and temperature change associated with a doubling of CO,-equivalent. In one analysis, Nordhaus (1991b) employs a linear damage function where climate changes gradually enough to permit social adjustment. In more recent work (Nordhaus, 1992, 1993), this assumption is altered so that damages increase quadratically with the deviation of mean global temperature from the pre-industrial norm. Here again the potential for “misplaced concreteness” (Daly and Cobb, 1989) comes to the fore: In the absence of reliable information, analysts must work with plausible yet arbitrary assumptions concerning the relationship between greenhouse gas emissions and the ensuing economic impacts.
4. Emissions
abatement:
measures
and costs
To stabilize current climatic conditions would require immediate emissions reductions of 50-80%
R.B. Howarrh, P.A. Monahan / Global and Planetary Change 1 I (1996) 187-199
for carbon dioxide, lo-20% for methane, 75-100% for CFCs, and 8O-85% for nitrous oxide (IPCC, 1991a). This level of abatement is not feasible in the short to intermedSate term given anticipated emissions levels and economic trends. There are, however, many possible abatement targets between the luissezfaire level of uncontrolled emissions and the draconian level of abatement required to achieve immediate climate stabilization. Non-energy secltor activities account for 54% of anticipated increases in radiative forcing: 9% from agriculture; 18% from land-use changes and deforestation; and 27% from industrial activities releasing CFCs, other halocarbons, and small amounts of CO, (IPCC, 1991a). R’educing these emissions carries unique difficulties and opportunities, particularly where subsistence activities are impacted. Mitigation costs for the agricultural sector are laden with ethical implications: Global food supplies both regionally and world-wide cannot be compromised for the sake of greenhouse gas emissions reduction. Some greenhouse gas mitigation activities can enhance agricultural output, but quantifying the costs associated with such strategies is currently beyond analytical capability. The cost of afforestation varies regionally and temporally. Afforestation may result in multiple benefits, with enhanced opportunities for indigenous peoples to sustainably utilize forest resources. Under the Montreal Protocol, CFCs are to be phased out over the next decade to protect the stratospheric ozone layer. Energy production and use are the single largest source of anthropogenic greenhouse gases emissions, with 46% of radiative forcing attributed to this sector in the 1980s. Relative to emissions from other sectors, energy-related emissions are the best understood and most readily quantified. Greenhouse gases released from energy transformation include CO, from fossil fuel combustion and CH, from coal mines and oil and gas facilities. Abatement options include switching from high to low carbon fuels, improving energy efficiency, and capturing emissions through carbon sequestration. While the costs of emissions abatement are relatively well established for industrialized nations, costs for developing countries are more speculative. Differential energy demand and economic growth patterns in the two regions imply the need for abatement strate-
191
gies tailored to local conditions. Two approaches have been employed to estimate the costs of reducing energy-related greenhouse gas emissions: economic models, which capture the response of the economy to changes in energy prices; and engineering studies, which examine the technological potential for emissions abatement. Economic models specify aggregate relationships between energy use and variables such as consumer income and industrial production, focusing on the response of energy use to price changes. These models suggest that an emissions tax could reduce U.S. carbon emissions in the year 2010 by 20% relative to 1990 levels at a direct cost of 0.9- 1.7% of national income (Weyant, 1993). If the revenues from the tax were used to reduce distortionaxy taxes on capital investment, the income losses associated with a carbon tax could be reduced by 35-100% (Shackleton et al., 1992). Engineering studies examine the potential for emissions abatement through the adoption of specific process technologies. A review by the U.S. National Academy of Sciences (1991) found that the implementation of technologies that are cost-effective at today’s energy prices could reduce the carbon intensity of the U.S. economy by some 40% over the long term. Given the slow turn-over of the capital stock, the achievement of this potential would take some time (approximately 10 years for most appliances and automobiles; 20 years for refrigerators; and 50 years for buildings). If this improvement could be achieved over the course of 20 years, however, the annual rate of change would come to some 2.5%/yr. Historical trends and forecasts of future developments point to improvements of only 0- 1%/yr in the absence of price changes or policy interventions. The interpretation of these results is a matter of some controversy. Schelling (1992, p. 1 l), for example, argues that technology analysts have been “unable to offer an explanation for why... low-cost or negative-cost technologies have not caught on”. Provisional analysis, however, suggests that market failures involving imperfect information and departures from substantive rationality may bias the market against energy-efficient technologies (Howarth and Winslow, 1994). Under this interpretation, regulatory measures such as stricter building codes and equipment performance standards might provide net COST
R.B. Howanh, PA Monahan/Global
192
savings while reducing the level of a carbon tax required to achieve specified emissions targets.
5. Normative criteria for climate policy The emergent facts of climate change are in themselves an insufficient basis for policy formulation. A normative framework is required to set those facts into perspective so that we may move from “is” to “ought”. Given the inherent uncertainties and intertemporal nature of climate change, a normative approach to this issue should be evaluated based on its treatment of intergenerational fairness and robustness in the face of pervasive uncertainty. Here we explore the ethical foundations of two alternatives: Economic efficiency, as operationalized using costbenefit analysis, and the principle of sustainable development (see also Brown, 1992; Broome, 1992).
6. Cost-benefit analysis and economic efficiency Cost-benefit analysis is derived from the principle that a policy change leads to an improvement in social welfare if it benefits at least some persons while leaving none worse off-a principle termed the &iciency criterion in welfare economics (Randall, 1987). If the monetary benefits accruing to winners are greater than the costs incurred by losers, then in principle the winners could compensate the losers so that the welfare of all individuals could be improved. The aggregation of costs and benefits that accrue at different points in time is a pervasive problem in the application of the efficiency criterion. In models of intertemporal equilibrium under perfect foresight, the after-tax rate of interest constitutes an appropriate measure of individuals’ marginal preference for consumption in sequential periods (Lind, 1982; Howarth and Norgaard, 1992). To express present and future benefits in comparable present-value units, net benefits that are realized I periods from the present are therefore discounted by the factor 1
a,=l-i,, 7=,
1 -rr7
(2)
and Planetary Change 11 (1996) 187-199
where r, is the interest or discount rate at date t and 6, = 0. Suppose that C, and B, are the flows of monetary costs and benefits realized at time f as a result of the proposed policy change. Then the net present value (NPV > of the net benefits yielded by the proposed policy change is given by NPV = i a,( B, - C,) 1-o
(3)
where the current date is normalized to r = 0 and T is the final date at which the policy has economic
impacts. If this measure is positive, then the policy change constitutes a potential efficiency improvement and is said to yield net positive benefits in the sense that the policy could be implemented along with appropriate income transfers so that all persons would be rendered better off. In general, a continuum of possible policies exists, and the efficiency criterion suggests the selection of the option that maximizes the NPV expression. In the case of climate change, an efficient policy regime may be identified by equating the marginal costs and benefits of greenhouse gas emissions, evaluated in present-value terms. 6.1. Applications
to climate policy
Cline (1992a, , 1992b) and Nordhaus (1993) use the damage and mitigation cost estimates discussed above to infer policy recommendations using the efficiency criterion. Nordhaus (1991b) applies partial equilibrium techniques to show that long-run emissions reductions ranging from near zero to about one-third relative to unconstrained levels are economically justified. He has since developed a more sophisticated intertemporal general equilibrium model that calls for emissions reductions of lo- 15% relative to a rising baseline (Nordhaus, 1992); this strategy yields relatively minor reductions in the rate at which greenhouse gases accumulate in the atmosphere. Nordhaus’ models employ an explicit optimization framework that maximizes the net benefits of greenhouse gas emissions abatement. The models, however, make no allowance for the possibility of catastrophic outcomes and hence presumably underestimate the efficient level of emissions reduction.
R.B. Howarth, PA. Monahun / Global and Planetary Change 1 I (1996) 187-199
Cline (1992a, , :l992b), in contrast, does not seek to identify an optimal level of emissions abatement but instead compares two alternative policies: a luissez faire case of .unconstrained emissions, and an aggressive abatement policy where worldwide carbon emissions are reduced to a permanent limit of 4 billion tonnes per year in comparison with baseline emissions of 6.7 billion tonnes in 1990 and 14 billion tonnes in 2050. Cline’s analysis is explicitly dynamic and allows for uncertainty through the specification of low-, mid-, and high-damage scenarios. He finds that the abatement policy is economically warranted provided that decision makers are risk averse, attaching a relatively high weight to unfavorable outcomes in the cost-benefit calculus. One source of disagreement between Cline and Nordhaus concerns the discounting procedures appropriate for evaluating long-term costs and benefits - a crucial issue in the economics of climate change. Cline (1992b), following recent developments in the theory of cost-benefit analysis, uses a social discount rate of lS%/yr, w:hich he argues corresponds to the after-tax return available on low-risk investments such as U.S. Treasury bills. The opportunity costs that arise when government policies crowd out high-yield private i:nvestments are addressed using a “shadow price of capital” approach. Nordhaus (1993, p. 22), in contrast, claims that Cline “arbitrarily assume[s] a near-zero discount rate” and argues for a discount rate on the order of 6-S% (Nordhaus, 1991b). While Cline’s general approach is well-grounded in established theory (Lind, 1982; Kolb and Scheraga, 1990), the empirical aspects of this question are perhaps worthy of further investigation. 6.2. Cost-benefit analysis and intergenerational
fair-
ness
All of the variables that go into a cost-benefit calculation-the cost of reducing greenhouse gas emissions, the associated environmental benefits, and discount rates-are reflections of anticipated economic conditions. The future path of the economy, however, is itself a. matter of collective choice; an optimal path should address questions of both intertemporal efficiency and intergenerational
193
fairness *. We could use the resources at our disposal to maximize our own gratification without regard to the welfare of future generations. Conversely, we might act to ensure that the life opportunities of future generations are no worse than our own. Either choice could be pursued with consummate economic efficiency, yet the efficient level of greenhouse gas emissions might vary sharply under the two scenarios. Suppose, for example, that climate impacts vary in linear proportion with world income. Then strong economic growth would raise the damages caused by greenhouse gas emissions relative to a low-growth scenario at each point in time. As we noted above, the discount rate appropriate for use in cost-benefit analysis is the after-tax return on capital investment. Economic growth is fueled by capital investment, with the rate of capital accumulation involving an equity decision on the level of wealth we wish to transfer to future generations. Increased accumulation implies a decrease in the marginal return on investment and hence a reduction in the social discount rate. Together, higher impacts and lower discount rates imply that it would be efficient to abate greenhouse gas emissions more aggressively in a high-growth world than in a low-growth alternative. This argument may be elaborated using formal economic models. Howarth and Norgaard (1992, , in press), for example, show that cost-benefit techniques may be used to identify efficient greenhouse gas emissions profiles in a hypothetical overlapping generations economy. The efficient outcome, however, depends on the degree of caring for the future. Fig. 1 shows the levels of key economic variablesper capita consumption, the capital stock, greenhouse gas concentrations, and the social discount rate-for two model runs. The “impoverished future” assumes that society cares little for posterity and thus depletes assets at its disposal. Under the “sustainable future”, in contrast, the present generation preserves capital goods and environmental quality for the sake of future generations. In each case, cost-be-
2Randall (1987) offers a good overview of the complementarity between efficiency and equity in the analysis of public policy. Howarth and Norgaard (1992, , in press) discuss the specific implications for long-term environmental management.
194
R.B. Howarth, P.A. Monahan/Global (a) Impoverished
Future
-
Con6umption
-
Capital Stock
+ Discount Rate
5
10
15
Generations
20
25
30
35
40
from Present
(b) Sustainable
Future
-
0
5
10
15
Generations
20
25
30
35
II (1996) 187-199
6.3. Cost-benefit analysis and interregional fairness
* Grsenhouse Gatss
0
and Planetary Chmge
Consumption
40
from Present
Fig. 1. Alternative future worlds. (a) Impoverished future. (b) Sustainable future.
nefit criteria are applied to identify an efficient greenhouse gas emissions profile. It is sometimes argued that technological progress will pave the way to a world of future abundance, obviating the need for worries about intergenerational fairness in the analysis and promulgation of public policy. But one cannot safely assume that conditions will improve in the future simply because they have improved in the past. Daly and Cobb (1989) argue that the trend towards economic progress has already reversed and that today’s young people will be unable to match the quality of life achieved by their parents in the absence of corrective policies. In the U.S., median incomes have been stagnant since the early 1970s with reductions for men under age 45 (Levy and Mumane, 1992). With this in mind, intergenerational fairness arises as a proper concern of decision makers that demands explicit attention in policy analysis.
Industrialized nations have historically benefitted from the use of technologies that generate greenhouse gas emissions and can arguably afford to take steps towards emissions abatement. Developing nations, in contrast, have contributed only modestly to the stock of greenhouse gases in the atmosphere; draconian emissions limitations might jeopardize the continued improvement of living conditions in areas where economic development is an urgent priority. Cost-benefit criteria suggest that emissions reductions should be realized wherever they can be achieved at minimum cost. Equity considerations, however, may demand cost-sharing through joint implementation or tradeable permit schemes to avoid imposing hardship on those already disadvantaged (Rose and Stevens, 1993). An analogous concern centers on the relative vulnerability of low-income nations lacking the means to adapt favorably to changing climatic conditions. In the extreme, unrestrained greenhouse gas emissions might take away some subsistence farmers’ most valuable asset - the stability and benevolence of nature. Reduced agricultural yields, accentuated drought-famine cycles, coastal inundation, and increases in the frequency and intensity of tropical storms might be compensated in whole or in part by development assistance and disaster relief programs, though responding after the fact may be more difficult and costly than proactive steps to limit the rate and magnitude of climate change. Limited attention has focused on the potential impacts of climate change in developing nations. Cline (1992b) and Nordhaus (1993), for example, seek to identify an efficient level of greenhouse gas emissions abatement for the world in toto based on the extrapolation of U.S. data. Even if the costs and benefits of climate change in developing nations were known with certainty, questions of interregional fairness would need to be assessed based on ethical principles that are logically distinct from the efficiency criterion. 6.4. Cost-benefit analysis and uncertainty Analyses of climate change often focus on central outcomes, averaging across low and high impact
R.B. Howarth, PA. Monahan/Global
scenarios to obtain an estimate of the most likely sequence of events. If individuals are risk averse, the standard principles of expected utility theory (see Kreps, 1990) sugge,st that this approach may provide a poor guide to policy given the substantial uncertainties associated with climate change. One approach to cost-benefit analysis under uncertainty is to use ad hoc procedures to adjust expected outcomes for risk. Since individuals demand relatively high expected rates of return on risky investments, cost-benefit analysts sometimes apply high discount rates in evaluating uncertain projects. Although tractable in practice, this approach raises theoretical objections (Wilson, 1982). A rational investor will demand a high expected return on an uncertain investment if its returns are positively correlated with the return on her/his overall investment portfolio. Conversely, she/he will accept comparatively low (or even negative) expected returns on assets that reduce risk by yielding high returns when the market as a whole turns sour. In principle, cost-benefit techniques may be used to identify social willingness to pay to reduce climatic risks (Manne and Richels, 1991). Suppose that there are i = O,..., n(r) possible outcomes or “states of nature” sZi at dates t = O,..., T. The probability of each state is Pr(s,,), and the costs and benefits of greenhouse gas emissions abatement are C,i and Bti. Then an efficient level of abatement may be identified by maximizing the net present-value expression T
n(t)
NPV = C C Pr( ~,i) S,i( B,i - C,i) t=O i=l
(4)
where the discount factor & varies across time and states of nature., reflecting individuals’ risk aversion and their relative well-being at sequential dates (Howarth and Norg,aard, in press). Exhaustive infonmation would be required to rigorously evaluate this criterion, which flows logically from the definition of intertemporal efficiency under uncertainty. The analyst would need to identify the complete set of possible future states, including their statistical probability, climatic conditions, and implications for human welfare. She would also need to gauge social preferences regarding time and risk, even for low-probability, high-consequence outcomes about which there is little objective information.
and Planetary Change I I (1996) 187-199
195
A heuristic version of this approach is offered by Cline (1992b), who examines the net benefits of greenhouse gas emissions abatement under three scenarios regarding the relationship between greenhouse gas stocks and climate damages: low, medium, and high. Each scenario is assigned a subjective weight based on its probability and the presumed risk aversion of the decision maker. Though such calculations are illustratively useful, they do not meet the demanding information requirements necessary to identify efficient resource allocations in the context of uncertainty. In this sense, cost-benefit analysis and the efficiency criterion are not fully operational in the analysis of climate policies. 6.5. The sustainability criterion Given the limitations of the efficiency criterion in addressing issues of uncertainty and intergenerational fairness, one might posit a social welfare function as a means of comparing and evaluating alternative strategies based on the comparative welfare of present and future generations across the complete range of possible outcomes. In principle, a social welfare function would simultaneously cope with questions of fairness and uncertainty by reducing social values to a single, well-defined criterion. In actuality, efforts to define an appropriate welfare function have failed for well-known theoretical and practical reasons (Sen, 1970). Even if a welfare function were fully identified, using it to compare alternative policies would run up against the same information requirements that confound cost-benefit analysis under uncertainty. We are left then to identify criteria that capture prevailing principles of intergenerational justice under uncertainty. The notion of intergenerational fairness as it is usually put forth in public debates over environmental policy takes the form of a constraint on the range of outcomes that are considered ethically permissible rather than a utilitarian definition of optimality (Howarth, in press). One widely accepted view is that economic and social development should be sustainable in the sense that it “meets the needs of the present without compromising the ability of future generations to meet their own needs” (WCED, 1987, p. 431, ensuring future generations the natural resources and environmental quality re-
196
R.B. Howarth, PA. Monahan/Global
quired to enjoy a favorable way of life. This approach is rooted in the moral premise that members of future generations should have life opportunities no worse than those of the present (Partridge, 1981; MacLean and Brown, 1983; Brown Weiss, 1989). Economists have interpreted sustainability as the technical requirement that the welfare of successive generations should be no lower than that of their predecessors (Pezzey, 1989). Because of the difficulties involved in defining the welfare of future generations, this approach is not directly operational. An alternative is to ensure that per capita consumption is nondecreasing over time. But aggregate economic indicators generally neglect non-market environmental amenities and the degradation and depletion of natural resource stocks (Repetto et al., 1989). Application of this approach would at a minimum require a careful reconsideration of conventional accounting techniques, and would likely run up against steep information requirements (Daly and Cobb, 1989). An alternative approach focuses on the conditions required to support a high standard of living into the indefinite future rather than the distribution of welfare across generations per se. Thus sustainability implies that we should ensure “the ability of future generations to meet their own needs” (WCED, 1987) or that future generations should inherit “an undiminished natural and economic endowment” (Brown et al., 1990). This approach does not require an exact definition of the welfare of future generations. It does, however, imply an obligation to conserve environmental quality for the benefit of future persons. Uncertainty concerning the future course of development constitutes a major concern of the sustainability criterion. How far should we go to protect future generations against the possibility of an inhospitable world? The composition of future generations will depend on the state of the world prevailing when they are born (Parfit, 1983). The individuals alive at a particular date under alternative contingent states should thus be regarded as ethically distinct potential generations, and sustainability would seem to require that the welfare of each potential generation be equal to or greater than that of its predecessor. Thus, in a world of uncertainty, the sustainability criterion may require sacrifices on the part of the present generation not only to raise the expected welfare of future generations but also to ensure that
and Planetary
Change
II (1996)
187-199
living standards are non-decreasing even under pessimistic circumstances. This is a strong supposition that needs to be placed in the context of competing social values. Few would argue, for example, that fifty percent of world income should be diverted to the construction of a planetary defense system to protect against the risk that future generations might be devastated by a collision between the Earth and a large asteroid. On the other hand, the world community has decided to incur significant costs to reduce the uncertain threat posed by ozone depletion in the upper atmosphere. Brown et al. (1990) argue that “A sustainable society is one that satisfies its needs without jeopardizing the prospects of future generations”. At a minimum, this interpretation suggests an obligation to reduce threats to future generations if so doing does not noticeably impact the subjective welfare of existing persons. 7. From normative criteria to climate policy Climate policy formulation involves balancing the interests of present and future generations and peoples of different world regions. A perceived benefit of climate stabilization is the reduction of risks to future generations, particularly those future persons most vulnerable to climate fluctuations because their poverty renders them susceptible to the vagaries of nature. Some analysts hold that the future impacts of climate change are too uncertain to justify the expenditure of present-day resources to mitigate greenhouse gas emissions. Gray and Rivkin (1991) argue for a “no-regrets” policy, whereby mitigation measures are implemented only if they yield clear benefits in ameliorating other social problems. A no regrets strategy would embrace the phase-out of CFCs under the Montreal Protocol given the chemicals’ role in destroying stratospheric ozone. Tropical deforestation is caused in part by policies that promote land-use changes through direct and indirect subsidies. Similarly, many nations subsidize the use of fossil fuels. A no-regrets strategy would reduce such subsidies to improve economic efficiency while at the same time slowing climate change. It would also embrace measures facilitating the adoption of cost-effective energy-efficient technologies.
R.B. Howarrh, PA Monahan/Global
The achievement of substantial long-term reductions in greenhouse gas emissions, however, implies moving beyond no-regrets measures to policies with costs ranging from ambiguous to rather large. Stabilization of today’s climate would require immediate carbon dioxide emissions reductions of some 60% relative to today’s level. Such reductions are arguably politically and socially infeasible. The question is thus not whether we should permit climate change or not, but what steps (if any> should be taken today to reduce the magnitude and timing of change. Krause et al. (19139)argue that human and ecological systems could adapt to temperatures 2.5”C above the pre-industrial norm if the rate of change was limited to O.l”C per decade. Based on this supposition, they identify maximum permissible greenhouse gas concentrations for each future date and work backwards to identify emissions constraints. They conclude that global emissions of carbon dioxide in industrialized nations should be reduced by 20% through 2005 with reductions of 50% and 75% achieved by 2015 and 2030. To support the sustained improvement of living standards, developing countries would be permitted to increase emissions by 50-100% over the short term, with emissions retuming to current levels by 2030. This approach is generally consistent with the United Nations Framework Convention on Climate Change, which calls for (Article 2): “[the] stabilization Iof greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the climate system. Such a level should be achieved within a time frame sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not threatened and to enable economic development to proceed in a sustainable manner”. The agreement mandates industrialized nations to return ‘‘anthropoge nit emissions of carbon dioxide and other greenhouse gases not controlled by the Montreal Protocol” to their 1990 levels by the year 2000 (Article 4, paragraph 2(b)). Reducing carbon emissions in the United States by 20% by the year 2010 would require the imposition of a carbon 1:ax of perhaps $100 per tonne (Howarth and Winslow, 1994). Such a tax would likely reduce economic activity by some O-2%, de-
and Planetary Change I1 (1996) 187-199
197
pending on its coordination with other policies 3. One could not credibly advance a carbon tax as a no-regrets approach to climate stabilization. The impacts of moderate emissions abatement on subjective human welfare, however, are somewhat ambiguous. Suppose that a 20% reduction in carbon dioxide emissions by the year 2010 would reduce economic activity by 1% in the industrialized world. Phased in gradually over time, this change would reduce economic growth by about O.O7%/yr. Except for workers in directly affected sectors such as coal mining and the supply of energy-efficient equipment, these changes might well pass unnoticed by most members of society. Perceived losses, where they occurred, could presumably be compensated through job retraining and other programs appropriate in a dynamic economy characterized by structural change. We have established that greenhouse gas emissions, left unmitigated, threaten uncertain but potentially serious impacts on the welfare of future generations. Catastrophic change is a distinct possibility that is consistent with prevailing knowledge of the interactions between climate and biogeochemical systems. There is good reason to believe that the costs of reducing greenhouse gas emissions in the industrialized nations are today zero or negative at the margin, while moderate abatement would impose modest costs with limited impacts on the subjective well-being of most individuals. To move from these premises to clear policy recommendations requires use of a corollary of the sustainability principle: Inhabitants of today’s world are morally obligated to take steps to reduce catastrophic risks to members of future generations if doing so would not noticeably diminish their own quality of life. This form of argument does not require a precise characterization of the impacts of climate change on future society, nor does it imply that the probabilities of extreme events must be calculated with confidence. Instead, we construct a simple two-part test based on the concept of operational thresholds. Is there a non-trivial risk of catastrophic change? Can
3Zero net abatement costs might possibly arise given optimistic assumptions concerning the use of carbon tax revenues to offset distortionaty taxes and the adoption of cost-effective energy-efflcient technologies. See section IV above.
198
R.B. Howarth, PA. Monahan /Global and Planetary Change II (1996) 187-199
we reduce that risk without unduly compromising our own well-being? The greenhouse gas emissions guidelines embodied in the Framework Convention would seem to pass both parts of this test. The sustainability principle also sheds light on the appropriate sharing of burdens between North and South. While the primacy of the sustainability rule is sometimes attacked on the grounds that we should take care of today’s poor before worrying about the future, principles of intergenerational justice flow logically from principles of justice between contemporaries (Malnes, 1990; Howarth, 1992)-the distinction between the two is based on a false premise. The Framework Convention holds that the costs of climate stabilization should be borne by those most able to manage them-the wealthy nations of the industrialized North. Indeed, climate stabilization arguably implies the transfer of technology and other assets from North to South to permit sustained development while limiting the growth of greenhouse gas emissions. A number of questions remain for further research. To argue that the sustainability criterion supports immediate steps towards greenhouse gas emissions abatement leaves open the question of the extent of emissions reduction that might be required over the longer term. The practical role of cost-benefit analysis in supporting climate policy is also worthy of further consideration. In principle, intertemporal efficiency and intergenerational fairness are complementary, not contradictory, criteria for resource allocation. In practice, economic analysis provides a means of modeling the impacts of policy measures on indicators such as employment and national income growth while offering concrete guidance as to how social objectives concerning environmental quality can be achieved at minimum social cost. References Broecker, W.S., 1987. Unpleasant surprises in the greenhouse. Nature, 328: 123- 126. Broome, J., 1992. Counting the Cost of Global Warming. White Horse Press, Cambridge. Brown, L.R., Flavin, C. and Postel, S., 1990. Picturing a sustainable society. In: Worldwatch Institute, State of the World 1990. Norton, New York, pp. 173-190.
Brown, P.G., 1992. Climate change and the planetary trust. Energy Policy, 20: 208-222. Brown Weiss, E., 1989. In Fairness to Future Generations: International Law, Common Patrimony, and Intergenerational Equity. Transnational, Dobbs Ferry, New York. Cline, W., 1992a. Global Warming: The Economic Stakes. Inst. Int. Econ., Washington. Cline, W., 1992b. The Economics of Global Warming. Inst. Int. Econ., Washington. Daly, H.E. and Cobb, J.B., 1989. For the Common Good: Redirecting the Economy Toward Community, the Environment, and a Sustainable Future. Beacon Press, Boston. Dansgaard, W., White, J.W.C. and Johnsen, S.J., 1989. The abrupt termination of the Younger Dryas climate event. Nature, 339: 532-534. Gleick, P.H., 1992. Effects of climate change on shared fresh water resources. In: I. Mintzer (Editor), Confronting Climate Change: Risks, Implications and Responses. Cambridge Univ. Press, pp. 127- 140. Gray, C.B. and Rivkin Jr., D.B., 1991. A “no regrets” environmental policy. Foreign Policy, N83: 47-64. Howarth, R.B., 1992. Intergenerational justice and the chain of obligation. Environ. Values, 1: 133- 140. Howarth, R.B., in press. Sustainability under uncertainty: a Kantian approach. Land Economics. Howarth, R.B. and Monahan, P.A., 1992. Economics, Ethics, and Climate Policy. Stockholm Environ. Inst., Stockholm. Howarth, R.B. and Norgaard, R.B., 1992. Environmental valuation under sustainable development. Am. Econ. Rev., 82: 473-477. Howarth, R.B. and Norgaard, R.B., in press. Intergenerational choices under global environmental change. In: D. Bromley (Editor), Handbook of Environmental Economics. Blackwell, Oxford. Howarth, R.B. and Winslow, M.A., 1994. Energy use and CO, emissions reduction: integrating pricing and regulatory policies. Energy, 19: 855-867. Hulm, P., 1989. A climate of crisis: global warming and the South Pacific islands. Assoc. South Pac. Environ. Inst., Port Moresby, Papua New Guinea. IPPC (Intergovernmental Panel on Climate Change), 1991a. Climate Change: The IPCC Response Strategies. Island Press, Washington. IPCC (Intergovernmental Panel on Climate Change), 1991b. Climate Change: The IPCC Scientific Assessment. Cambridge Univ. Press. Kolb, J.A. and Scheraga, J.D., 1990. Discounting the benefits and costs of environmental regulations. J. Policy Analys. Manage., 9: 381-390. Krause, F., Bach, W. and Koomey, J., 1989. Energy Policy in the Greenhouse, Volume I. Int. Project Sustainable Energy Paths, El Cenito, CA. Kreps, D.M., 1990. A Course in Microeconomic Theory. Princeton Univ. Press. Levy, F. and Mumane, R.J., 1992. U.S. earnings levels and earnings inequality: a review of recent trends and proposed explanations. J. Econ. Literature, 30: 1333-1381.
R.B. Howarth, PA. Monahan/Global Lewis, J., 1989. Sea level rise: some implications for Tuvalu. Ambio, 18: 458-459. Lind, R., 1982. Discounting for Time and Risk in Energy Policy. Resour. Future. MacLean, D. and Brown, PG., 1983. Energy and the Future. Rowman and Littlefield, Totawa, NJ. Malnes, R., 1990. The Environment and Duties to Future Generations: An Elaboration of Sustainable Development. Fridtjof Nansen Inst., Lysaker. Manne, A.S. and RicheH, R.G., 1991. Buying greenhouse insurance. Energy Policy, 6: 543-552. Mearns, L.O., Katz, R.W. and Schneider, S., 1984. Extreme high temperature events: changes in their probabilities with changes in mean temperatures. J. Clim. Appl. Meteorol., 23: 16011613. Nordhaus, W.D., 1989. The economics of the greenhouse effect. Paper presented to the MIT Workshop on Energy and Environmental Modeling and Policy Analysis. Nordhaus, W.D., 1990. Economic policy in the face of global warming. Dep. Econ., Yale Univ. Nordhaus, W.D., 1991a. Economic approaches to global warming. In: R. Dombusch and J. Poterba (Editors), Global Warming: Economic Policy Responses. MIT Press, Cambridge, MA. Nordhaus, W.D., 1991b. To slow or not to slow: the economics of the greenhouse effect. Econ. J., 101: 920-937. Nordhaus, W.D., 1992. An optimal transition path for controlling greenhouse gases. Science, 258: 13 15- 13 19. Nordhaus, W.D., 1993. Reflections on the economics of climate change. J. Econ. Perspcct., 7: 1 l-25. Oeschger, H. and Mintzer, I., 1992. Lessons from the ice cores: rapid climate changes during the last 160,000 years. In: I. Mintzer (Editor), Confronting Climate Change: Risks, Implications and Responses. Cambridge Univ. Press, pp. 55-64. Parfit, D., 1983. Energy policy and the further future: the identity problem. In: D. MacLean and P.G. Brown (Editors), Energy and the Future. Rowman and Littlefield, Totawa, NJ, pp. 38-58. Parry, M.L., 1990. Climate Change and World Agriculture. Earthscan, London. Partridge, E., 1981. Responsibilities to Future Generations. Prometheus, Buffalo. Pezzey, J., 1989. Economic Analysis of Sustainable Growth and Sustainable Development. World Bank, Washington. Randall, A., 1987. Resource Economics: An Economic Approach to Natural Resource and Environmental Policy. Wiley, New York, 2nd ed. Repetto, R., Magrath, PI., Wells, M., Beer, C. and Rossini, F., 1989. Wasting Assets: Natural Resources in the National Income Accounts. World Resour. Inst., Washington.
and Planetary Change II (1996) 187-199
199
Revelle, R., 1983. Methane hydrates in continental slope sediments and increasing atmospheric carbon dioxide. In: National Research Council, Changing Climate. Natl. Acad. Press, Washington, pp. 252-261. Rind, D., Goldberg, R., Hansen, J., Rosenzweig, C. and Ruedy, R., 1990. Potential evapotranspiration and the likelihood of future drought. J. Geophys. Res., 95: 9983-10,004. Rose, A. and Stevens, B., 1993. The efficiency and equity of marketable permits for CO, emissions. Resource Energy Econ., 15: 117-146. Schelling, T.C., 1992. Some economics of global warming. Am. Econ. Rev., 82: 1-14. Schimel, D., 1990. Biogeochemical feedbacks in the earth system, In: J. Leggett (Editor), Global Warming: The Greenpeace Report. Oxford Univ. Press. Schneider, S., 1989. Global Warming. Sierra Club Books, San Francisco. Sen, A.K., 1970. Collective Choice and Social Welfare. HoldenDay, San Francisco. Shackleton, R., Shelby, M., Cristofaro, A., Brinner, R., Yanchar, J., Goulder, L., Jorgenson, D., Wilcoxen, P. and Pauly, P., 1992. The efficiency value of carbon tax revenues. U.S. Environ. Protect. Agency, Washington. Svedin, U. and Aniansson, B., 1987. Surprising Futures. Swedish Count. Plann. Coord. Res., Stockholm. USEPA (U.S. Environmental Protection Agency), 1989a. Policy Options for Stabilizing Global Climate. Government Printing Office, Washington. USEPA (U.S. Environmental Protection Agency), 1989b. The Potential Effects of Global Climate Change on the United States. Government Printing Oftice, Washington. USEPA (U.S. Environmental Protection Agency), 1990. The Potential Effects of Global Climate Change on the United States. Hemisphere, Washington. U.S. National Academy of Sciences, 1991. Policy Implications of Greenhouse Warming: Report of the Mitigation Panel. Natl. Acad. Press, Washington. Weyant, J.P., 1993. Costs of reducing global carbon emissions. J. Econ. Perspect., 7: 27-46. Wilson, R., 1982. Risk measurement of public projects. In: R. Lind (Editor), Discounting for Time and Risk in Energy Policy. Resour. Future, pp. 205-249. Woodwell, G.M., 1990. The effects of global warming. In: J. Leggett (Editor), Global Warming: The Greenpeace Report. Oxford Univ. Press, New York, pp. 116- 132. WCED (World Commission on Environment and Development), 1987. Our Common Future. Oxford Univ. Press, New York.
ELSEVIER
Global and Planetary Change
11( 1996) 187- 199
Economics, ethics, and climate policy: framing the debate
’
Richard B. Howarth a, Patricia A. Monahan b a Environmental Studies Program, University of California, Santa Crut, CA 95064, USA b Energy and Environment Division, Lawrence Berkeley Laboratory, Berkeley, CA 94720, USA Accepted
30
November 1995
Abstract This paper examines the economic and ethical dimensions of climate policy in light of existing knowledge of the impacts of global warming and the costs of greenhouse gas emissions abatement. We find that the criterion of economic efficiency, operationalized through cost-benefit analysis, is ill-equipped to cope with the pervasive uncertainties and issues of intergenerational fairness that characterize climate change. In contrast, the concept of sustainable development-that today’s policies should ensure that future generations enjoy life opportunities undiminished relative to the present-is a normative criterion that explicitly addresses the uncertainties and distributional aspects of global environmental change. If one interprets the sustainability criterion to imply that it is morally wrong to impose catastrophic risks on unborn generations when reducing those risks would not noticeably diminish the quality of life of existing persons, a case can be made for significant steps to reduce greenhouse gas emissions.
1. Introduction Debates over the appropriate policy response to global climate change involve fundamental disputes concerning social values and the role of science in policy analysis. Some argue that policies should equate the marginal costs and benefits of greenhouse gas emissions as defined in economic terms. Others hold that climate change involves ethical dimensions that are not easily incorporated in economic analysis -in particular, that future generations have a moral right to a stable climate that imposes binding obligations on present actions.
’ This article is a revised and much-condensed version of the report “Economics, Ethics, and Climate Policy” (Howarth and Monahan, 1992). We thank the Stockholm Environment Institute for fmancial support. 0921-8181/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved SSDI 0921-8181(95)00052-6
If we were certain that the impacts of climate change would be gradual and easily managed by social adaptation, reducing greenhouse gas emissions would hardly appear as an urgent policy priority. If, on the other hand, we were certain that unmitigated emissions would imply catastrophe for the well-being of future societies, few would argue for a “go slow” approach to policy intervention. But although scientists have reached broad agreement on a number of issues relating to climate change, a key fact is that both the dimensions of climate change and its impacts on human and ecological systems will remain highly uncertain. Coping with uncertainty is therefore a central policy challenge. This paper examines alternative frameworks for analyzing climate change as a global policy challenge with a special focus on cost-benefit analysis and principles of intergenerational fairness. We find that the criterion of economic efficiency as opera-
188
R.B. Howarth, PA. Monahan/Glohal
tionalized through cost-benefit analysis is poorly suited to the evaluation of climate policy for two fundamental reasons: (1) cost-benefit techniques are difficult to operationalize given the gross uncertainties surrounding climate change and its potential effects on future society; and (2) the approach is insensitive to the distribution of impacts between present and future generations, yet climate change may have pronounced implications for future economic conditions and environmental quality. Under the rubric of “sustainable development”, intergenerational fairness has been widely embraced as a criterion in environmental planning. We explore the ethical basis of the sustainability approach, its translation into operational criteria, and its implications for climate policy. The sustainability criterion may be taken to imply that it is morally wrong to impose catastrophic risks on members of unborn generations if reducing those risks would not significantly diminish the quality of life of existing persons. Under this interpretation, a case can be established for significant greenhouse gas emissions abatement given current knowledge of mitigation costs and the risks of climate change. Section II sketches the key facts and uncertainties of climate science as they relate to policy formulation. Section III outlines the potential impacts of climate change on human and environmental systems and the difficulties of translating those impacts into economic terms. Section IV examines the cost of policies to abate greenhouse gas emissions, while section V explores the logical foundations of costbenefit analysis and the principle of sustainable development as guides to public policy. We conclude by reviewing the principal findings that stem from our analysis.
2. Climate science: the emergent facts A working consensus has been reached regarding key aspects of climate science. Atmospheric concentrations of greenhouse gases [carbon dioxide, chlorofluorocarbons (CFCs), methane, nitrous oxide, and water vapor] have increased rapidly due to human activities such as fossil fuel combustion, refrigeration, air conditioning, deforestation, and agriculture. In the absence of steps to reduce greenhouse gas
and Planetary Change 11 (1996) 187-199
emissions, a doubling of atmospheric CO,-equivalent relative to the pre-industrial level is anticipated by 2025 (IPCC, 1991a). General circulation models (GCMS) project that a doubling of CO,-equivalent would raise the Earth’s mean temperature by 1.9-5.2”C, with 2.5”C as the best estimate (IPCC, 1991b). While such changes may seem small when compared against daily and seasonal variations, a change of only 1°C heralded the Little Ice Age between the 14th and 17th Century, inducing frequent crop failures and sporadic freezing of the Baltic Sea (Oeschger and Mintzer, 1992). A 5°C temperature increase would move the Earth to a climatic regime not experienced in over a million years. Beyond these essentials, scientific research points to a range of potential impacts that climate change might entail, from enhanced agricultural yields to the inundation of island nations. Given the prevailing state of knowledge, neither the probability nor severity of climate impacts can be gauged with confidence. There is, however, general agreement in the scientific community that the potential for catastrophic change exists. This section outlines some of the more likely impacts of climate change, as well as a number of catastrophic outcomes that might result from feedback processes that are not well understood by existing science. As temperatures warm over the course of the next century, sea level is expected to rise by up to one meter. Since half of humanity currently occupies coastal zones, a change of this magnitude would have far-reaching effects. Most at risk are poorer nations with densely populated coastal areas. Indonesia, with 15% of the world’s coastlines, is projected to lose 40% of its land surface (Schneider, 1989). A one-meter rise could inundate 15% of Bangladesh and all of the Republic of Maldives, Kiribati, the Marshall Islands, Tokelau, Tuvalu, and the Torres Strait Islands (Hulm, 1989; Lewis, 1989). Rapid climate change would alter the composition of ecosystems, with some species benefiting and others unable to migrate or adapt at the rate necessary for survival. Given a warming of O.l-1°C per decade, Woodwell (1990) finds that the capacity of natural communities to migrate is exceeded by factors of 100-1000 or more; a substantial loss of biodiversity is a distinct possibility. Impacts on hu-
R.B. Howarth. P.A. Monahan/Global
man systems will likely be felt most acutely in regions already under stress, such as those exposed to the natural hazards of coastal and riparian flooding, drought, land-slides, severe storms, or tropical cyclones. Climate change might also increase the frequency and intensity of tropical storms; diminish water supplies in areas already experiencing water shortages (Gleick, 1992); force the mass migration of human populations; and reduce agricultural yields, particularly in low-income regions unable to adapt to changing climatic c’onditions. Today’s GCMs do not fully incorporate feedback processes between atmospheric, oceanic, and biogeochemical systems. But as Schimel (1990, p. 68) points out, “Feedbacks between atmosphere and biosphere are non-linear, sensitive to initial conditions, and capable of enormous amplifications. Complex feedbacks in the Earth system can produce unexpected and potent responses... Without crying wolf, it is worthy of our concern as a society that biogeochemical and ecological feedbacks may result in more rapid environmental change than is predicted by purely physical models.” Some 10,000 years ago, a shut-down of ocean circulation may have contributed to regional temperature increases as large as 7°C and a 50% increase in rainfall over the course of a few decades (Dansgaard et al., 1989). Broecker (1987) suggests that changes in temperature and precipitation patterns may alter the circulation of the world’s oceans, inducing unpredictable but potentially catastrophic feedbacks. A second potential feedback is the accentuation of greenhouse gas emissions from the biosphere. Global warming could accelerate the anaerobic decay of organic matter, stimulating methane generation, as well as melting methane clathrates stored in Arctic sediments. Revelle (1983) estimates that clathrate liberation might amplify global warming by 0.6518°C. Although such results point to the potential for catastrophic change as a key fact of climate science, neither the probability nor consequences of such outcomes is easily quantified. Yet as Svedin and Aniansson (1987) warn, by “leaving out the external shocks, nonlinear responses, and discontinuous behavior so typical of social and natural systems, sur-
and Planetary Change 11 (1996) 187-199
189
prise-free analysis leaves us unprepared to interpret a host of not-improbable eventualities”.
3. The economic costs of climate change The application of cost-benefit analysis to the assessment of climate policy requires monetary estimates of the costs and benefits of climate change. In principle, monetizing costs and benefits would allow for the comparison of dissimilar impacts in a consistent framework to identify an efficient level of greenhouse gas emissions abatement. The rigorous application of cost-benefit analysis, however, requires information on the full range of possible outcomes and their respective probabilities. Since neither can be identified with confidence given the prevailing state of scientific knowledge, economic analysis is unable to provide unambiguous estimates of the relevant costs and benefits. Using the same base of scientific information, Nordhaus ( 19891, Nordhaus, 1990, 199 1a> and Cline (1992a) Cline, 1992b) derive rather different estimates of the damages a doubling of atmospheric CO,-equivalent would impose on the U.S. economy (Table 1). A comparison of these studies lends insight into how analysts’ worldviews and intuitions enter the valuation calculus. Agricultural impacts reflect the greatest divergence between the Nordhaus and Cline damage estimates. While elevated CO, concentrations and longer growing seasons enhance plant yields under laboratory conditions, shifts in soil moisture (Parry, 1990), increased range of pests and diseases (USEPA, 1990), and changes in the probability of extreme weather events (Meams et al., 1984) might lead to reduced yields. Nordhaus expects the costs and benefits to agriculture to balance at $10 billion each with zero net costs. This figure reflects a USEPA (1989a, , 1989b) estimate that assumes a greater carbon fertilization effect than will be realized with a doubling of CO,-equiualent. Cline tempers USEPA’s estimate to account for a lower carbon fertilization effect and includes the costs from increased incidence of severe drought (Rind et al., 19901, estimating net costs of $8.6 billion for the 1981 breakdown of the economy. Nordhaus and Cline also differ in their approach to nonmarket impacts such as species extinction and
190
R.B. Howarth, P.A. Monahan/Global
Table 1 Damage costs to the U.S. economy for a doubling of CO,-equivatent (billion 1981$ per year) Study
Cline, 1992a a
Nordhaus, 1991a
1981 National income Forestry Electricity/Energy Air conditioning Heating Sea level rise Loss of land Coastal protection Recreation Ozone impacts Health Water supply Agriculture Species loss Migration Hurricanes Urban infrastructure Total of 1981 income
2415.1 1.6
2415.1 small
5.5 -0.6
1.65 - 1.16
2.9 0.6 0.8 1.7 2.9 3.5 8.6 2.0 0.2 0.4 0.05 30.1 1.25%
1.5 3.7 small not quantified not quantified not quantified -9.7-10.6 not quantified not quantifted not quantified not quantified 6.2% 0.25%
a Cline’s initial estimates were based on 1990 prices and economic conditions. We multiplied these figures by the ratio of nominal economic activity in 1981 and 1990 to obtain the estimates reported here.
loss of human life in natural disasters. Nordhaus (1991a, p. 44) acknowledges that nonmarket goods “escape the net of the national income accounts and might affect the calculations” but notes that although “[some] people will place a high moral, aesthetic, or environmental value on preventing climate change,... I know of no serious estimates of what people are willing to pay to stop greenhouse warming”. Cline, in contrast, explicitly examines the potential costs of impacts on recreation, water supply, species loss, migration, hurricane damage, urban infrastructure, health and welfare, and forestry. He finds these costs to be cumulatively greater than the costs of climate change on agricultural output. Despite these differences, Cline and Nordhaus work with closely similar damage cost estimates (1 .O vs. 1.25% of economic activity for a doubling of CO,-equivalent) when integrating the costs and benefits of greenhouse gas emissions abatement. Although Nordhaus’ direct cost estimate is only onefifth of Cline’s figure, he judgmentally adjusts his cost coefficient from 0.25 to 1.0% to account for
und Planetmy
Chunge II (1996) 187-199
omitted factors (Nordhaus, 1992). Both Cline’s and Nordhaus’ cost estimates are plausible based on existing science, yet each relies strongly on subjective judgements that are, at present, not amenable to empirical test. Much larger impacts might arise if climate change turns out to be worse than expected; smaller impacts might also occur. Since the key uncertainties are not quantified, such estimates provide an insufficient basis for the rigorous application of cost-benefit techniques. A final issue concerns the functional relationship between greenhouse gas concentrations and related economic damages. The literature focuses mainly on the expected impacts of a doubling of atmospheric CO,-equivalent, yet the potential impacts of much larger concentrations are of key policy relevance. Cline (1992b, p. 376) employs a geometric damage function under the premise that the “greenhouse effect poses major risks, especially over the very long term of two to three centuries, by which time temperatures could rise by as much as 10” to 18°C”. He specifies climate damages per unit of economic activity (d) as d= d,, x ( AT/ATo)1.3
(1)
where AT is the temperature increase caused by climate change while d, and AT, are the damage level and temperature change associated with a doubling of CO,-equivalent. In one analysis, Nordhaus (1991b) employs a linear damage function where climate changes gradually enough to permit social adjustment. In more recent work (Nordhaus, 1992, 1993), this assumption is altered so that damages increase quadratically with the deviation of mean global temperature from the pre-industrial norm. Here again the potential for “misplaced concreteness” (Daly and Cobb, 1989) comes to the fore: In the absence of reliable information, analysts must work with plausible yet arbitrary assumptions concerning the relationship between greenhouse gas emissions and the ensuing economic impacts.
4. Emissions
abatement:
measures
and costs
To stabilize current climatic conditions would require immediate emissions reductions of 50-80%
R.B. Howarrh, P.A. Monahan / Global and Planetary Change 1 I (1996) 187-199
for carbon dioxide, lo-20% for methane, 75-100% for CFCs, and 8O-85% for nitrous oxide (IPCC, 1991a). This level of abatement is not feasible in the short to intermedSate term given anticipated emissions levels and economic trends. There are, however, many possible abatement targets between the luissezfaire level of uncontrolled emissions and the draconian level of abatement required to achieve immediate climate stabilization. Non-energy secltor activities account for 54% of anticipated increases in radiative forcing: 9% from agriculture; 18% from land-use changes and deforestation; and 27% from industrial activities releasing CFCs, other halocarbons, and small amounts of CO, (IPCC, 1991a). R’educing these emissions carries unique difficulties and opportunities, particularly where subsistence activities are impacted. Mitigation costs for the agricultural sector are laden with ethical implications: Global food supplies both regionally and world-wide cannot be compromised for the sake of greenhouse gas emissions reduction. Some greenhouse gas mitigation activities can enhance agricultural output, but quantifying the costs associated with such strategies is currently beyond analytical capability. The cost of afforestation varies regionally and temporally. Afforestation may result in multiple benefits, with enhanced opportunities for indigenous peoples to sustainably utilize forest resources. Under the Montreal Protocol, CFCs are to be phased out over the next decade to protect the stratospheric ozone layer. Energy production and use are the single largest source of anthropogenic greenhouse gases emissions, with 46% of radiative forcing attributed to this sector in the 1980s. Relative to emissions from other sectors, energy-related emissions are the best understood and most readily quantified. Greenhouse gases released from energy transformation include CO, from fossil fuel combustion and CH, from coal mines and oil and gas facilities. Abatement options include switching from high to low carbon fuels, improving energy efficiency, and capturing emissions through carbon sequestration. While the costs of emissions abatement are relatively well established for industrialized nations, costs for developing countries are more speculative. Differential energy demand and economic growth patterns in the two regions imply the need for abatement strate-
191
gies tailored to local conditions. Two approaches have been employed to estimate the costs of reducing energy-related greenhouse gas emissions: economic models, which capture the response of the economy to changes in energy prices; and engineering studies, which examine the technological potential for emissions abatement. Economic models specify aggregate relationships between energy use and variables such as consumer income and industrial production, focusing on the response of energy use to price changes. These models suggest that an emissions tax could reduce U.S. carbon emissions in the year 2010 by 20% relative to 1990 levels at a direct cost of 0.9- 1.7% of national income (Weyant, 1993). If the revenues from the tax were used to reduce distortionaxy taxes on capital investment, the income losses associated with a carbon tax could be reduced by 35-100% (Shackleton et al., 1992). Engineering studies examine the potential for emissions abatement through the adoption of specific process technologies. A review by the U.S. National Academy of Sciences (1991) found that the implementation of technologies that are cost-effective at today’s energy prices could reduce the carbon intensity of the U.S. economy by some 40% over the long term. Given the slow turn-over of the capital stock, the achievement of this potential would take some time (approximately 10 years for most appliances and automobiles; 20 years for refrigerators; and 50 years for buildings). If this improvement could be achieved over the course of 20 years, however, the annual rate of change would come to some 2.5%/yr. Historical trends and forecasts of future developments point to improvements of only 0- 1%/yr in the absence of price changes or policy interventions. The interpretation of these results is a matter of some controversy. Schelling (1992, p. 1 l), for example, argues that technology analysts have been “unable to offer an explanation for why... low-cost or negative-cost technologies have not caught on”. Provisional analysis, however, suggests that market failures involving imperfect information and departures from substantive rationality may bias the market against energy-efficient technologies (Howarth and Winslow, 1994). Under this interpretation, regulatory measures such as stricter building codes and equipment performance standards might provide net COST
R.B. Howanh, PA Monahan/Global
192
savings while reducing the level of a carbon tax required to achieve specified emissions targets.
5. Normative criteria for climate policy The emergent facts of climate change are in themselves an insufficient basis for policy formulation. A normative framework is required to set those facts into perspective so that we may move from “is” to “ought”. Given the inherent uncertainties and intertemporal nature of climate change, a normative approach to this issue should be evaluated based on its treatment of intergenerational fairness and robustness in the face of pervasive uncertainty. Here we explore the ethical foundations of two alternatives: Economic efficiency, as operationalized using costbenefit analysis, and the principle of sustainable development (see also Brown, 1992; Broome, 1992).
6. Cost-benefit analysis and economic efficiency Cost-benefit analysis is derived from the principle that a policy change leads to an improvement in social welfare if it benefits at least some persons while leaving none worse off-a principle termed the &iciency criterion in welfare economics (Randall, 1987). If the monetary benefits accruing to winners are greater than the costs incurred by losers, then in principle the winners could compensate the losers so that the welfare of all individuals could be improved. The aggregation of costs and benefits that accrue at different points in time is a pervasive problem in the application of the efficiency criterion. In models of intertemporal equilibrium under perfect foresight, the after-tax rate of interest constitutes an appropriate measure of individuals’ marginal preference for consumption in sequential periods (Lind, 1982; Howarth and Norgaard, 1992). To express present and future benefits in comparable present-value units, net benefits that are realized I periods from the present are therefore discounted by the factor 1
a,=l-i,, 7=,
1 -rr7
(2)
and Planetary Change 11 (1996) 187-199
where r, is the interest or discount rate at date t and 6, = 0. Suppose that C, and B, are the flows of monetary costs and benefits realized at time f as a result of the proposed policy change. Then the net present value (NPV > of the net benefits yielded by the proposed policy change is given by NPV = i a,( B, - C,) 1-o
(3)
where the current date is normalized to r = 0 and T is the final date at which the policy has economic
impacts. If this measure is positive, then the policy change constitutes a potential efficiency improvement and is said to yield net positive benefits in the sense that the policy could be implemented along with appropriate income transfers so that all persons would be rendered better off. In general, a continuum of possible policies exists, and the efficiency criterion suggests the selection of the option that maximizes the NPV expression. In the case of climate change, an efficient policy regime may be identified by equating the marginal costs and benefits of greenhouse gas emissions, evaluated in present-value terms. 6.1. Applications
to climate policy
Cline (1992a, , 1992b) and Nordhaus (1993) use the damage and mitigation cost estimates discussed above to infer policy recommendations using the efficiency criterion. Nordhaus (1991b) applies partial equilibrium techniques to show that long-run emissions reductions ranging from near zero to about one-third relative to unconstrained levels are economically justified. He has since developed a more sophisticated intertemporal general equilibrium model that calls for emissions reductions of lo- 15% relative to a rising baseline (Nordhaus, 1992); this strategy yields relatively minor reductions in the rate at which greenhouse gases accumulate in the atmosphere. Nordhaus’ models employ an explicit optimization framework that maximizes the net benefits of greenhouse gas emissions abatement. The models, however, make no allowance for the possibility of catastrophic outcomes and hence presumably underestimate the efficient level of emissions reduction.
R.B. Howarth, PA. Monahun / Global and Planetary Change 1 I (1996) 187-199
Cline (1992a, , :l992b), in contrast, does not seek to identify an optimal level of emissions abatement but instead compares two alternative policies: a luissez faire case of .unconstrained emissions, and an aggressive abatement policy where worldwide carbon emissions are reduced to a permanent limit of 4 billion tonnes per year in comparison with baseline emissions of 6.7 billion tonnes in 1990 and 14 billion tonnes in 2050. Cline’s analysis is explicitly dynamic and allows for uncertainty through the specification of low-, mid-, and high-damage scenarios. He finds that the abatement policy is economically warranted provided that decision makers are risk averse, attaching a relatively high weight to unfavorable outcomes in the cost-benefit calculus. One source of disagreement between Cline and Nordhaus concerns the discounting procedures appropriate for evaluating long-term costs and benefits - a crucial issue in the economics of climate change. Cline (1992b), following recent developments in the theory of cost-benefit analysis, uses a social discount rate of lS%/yr, w:hich he argues corresponds to the after-tax return available on low-risk investments such as U.S. Treasury bills. The opportunity costs that arise when government policies crowd out high-yield private i:nvestments are addressed using a “shadow price of capital” approach. Nordhaus (1993, p. 22), in contrast, claims that Cline “arbitrarily assume[s] a near-zero discount rate” and argues for a discount rate on the order of 6-S% (Nordhaus, 1991b). While Cline’s general approach is well-grounded in established theory (Lind, 1982; Kolb and Scheraga, 1990), the empirical aspects of this question are perhaps worthy of further investigation. 6.2. Cost-benefit analysis and intergenerational
fair-
ness
All of the variables that go into a cost-benefit calculation-the cost of reducing greenhouse gas emissions, the associated environmental benefits, and discount rates-are reflections of anticipated economic conditions. The future path of the economy, however, is itself a. matter of collective choice; an optimal path should address questions of both intertemporal efficiency and intergenerational
193
fairness *. We could use the resources at our disposal to maximize our own gratification without regard to the welfare of future generations. Conversely, we might act to ensure that the life opportunities of future generations are no worse than our own. Either choice could be pursued with consummate economic efficiency, yet the efficient level of greenhouse gas emissions might vary sharply under the two scenarios. Suppose, for example, that climate impacts vary in linear proportion with world income. Then strong economic growth would raise the damages caused by greenhouse gas emissions relative to a low-growth scenario at each point in time. As we noted above, the discount rate appropriate for use in cost-benefit analysis is the after-tax return on capital investment. Economic growth is fueled by capital investment, with the rate of capital accumulation involving an equity decision on the level of wealth we wish to transfer to future generations. Increased accumulation implies a decrease in the marginal return on investment and hence a reduction in the social discount rate. Together, higher impacts and lower discount rates imply that it would be efficient to abate greenhouse gas emissions more aggressively in a high-growth world than in a low-growth alternative. This argument may be elaborated using formal economic models. Howarth and Norgaard (1992, , in press), for example, show that cost-benefit techniques may be used to identify efficient greenhouse gas emissions profiles in a hypothetical overlapping generations economy. The efficient outcome, however, depends on the degree of caring for the future. Fig. 1 shows the levels of key economic variablesper capita consumption, the capital stock, greenhouse gas concentrations, and the social discount rate-for two model runs. The “impoverished future” assumes that society cares little for posterity and thus depletes assets at its disposal. Under the “sustainable future”, in contrast, the present generation preserves capital goods and environmental quality for the sake of future generations. In each case, cost-be-
2Randall (1987) offers a good overview of the complementarity between efficiency and equity in the analysis of public policy. Howarth and Norgaard (1992, , in press) discuss the specific implications for long-term environmental management.
194
R.B. Howarth, P.A. Monahan/Global (a) Impoverished
Future
-
Con6umption
-
Capital Stock
+ Discount Rate
5
10
15
Generations
20
25
30
35
40
from Present
(b) Sustainable
Future
-
0
5
10
15
Generations
20
25
30
35
II (1996) 187-199
6.3. Cost-benefit analysis and interregional fairness
* Grsenhouse Gatss
0
and Planetary Chmge
Consumption
40
from Present
Fig. 1. Alternative future worlds. (a) Impoverished future. (b) Sustainable future.
nefit criteria are applied to identify an efficient greenhouse gas emissions profile. It is sometimes argued that technological progress will pave the way to a world of future abundance, obviating the need for worries about intergenerational fairness in the analysis and promulgation of public policy. But one cannot safely assume that conditions will improve in the future simply because they have improved in the past. Daly and Cobb (1989) argue that the trend towards economic progress has already reversed and that today’s young people will be unable to match the quality of life achieved by their parents in the absence of corrective policies. In the U.S., median incomes have been stagnant since the early 1970s with reductions for men under age 45 (Levy and Mumane, 1992). With this in mind, intergenerational fairness arises as a proper concern of decision makers that demands explicit attention in policy analysis.
Industrialized nations have historically benefitted from the use of technologies that generate greenhouse gas emissions and can arguably afford to take steps towards emissions abatement. Developing nations, in contrast, have contributed only modestly to the stock of greenhouse gases in the atmosphere; draconian emissions limitations might jeopardize the continued improvement of living conditions in areas where economic development is an urgent priority. Cost-benefit criteria suggest that emissions reductions should be realized wherever they can be achieved at minimum cost. Equity considerations, however, may demand cost-sharing through joint implementation or tradeable permit schemes to avoid imposing hardship on those already disadvantaged (Rose and Stevens, 1993). An analogous concern centers on the relative vulnerability of low-income nations lacking the means to adapt favorably to changing climatic conditions. In the extreme, unrestrained greenhouse gas emissions might take away some subsistence farmers’ most valuable asset - the stability and benevolence of nature. Reduced agricultural yields, accentuated drought-famine cycles, coastal inundation, and increases in the frequency and intensity of tropical storms might be compensated in whole or in part by development assistance and disaster relief programs, though responding after the fact may be more difficult and costly than proactive steps to limit the rate and magnitude of climate change. Limited attention has focused on the potential impacts of climate change in developing nations. Cline (1992b) and Nordhaus (1993), for example, seek to identify an efficient level of greenhouse gas emissions abatement for the world in toto based on the extrapolation of U.S. data. Even if the costs and benefits of climate change in developing nations were known with certainty, questions of interregional fairness would need to be assessed based on ethical principles that are logically distinct from the efficiency criterion. 6.4. Cost-benefit analysis and uncertainty Analyses of climate change often focus on central outcomes, averaging across low and high impact
R.B. Howarth, PA. Monahan/Global
scenarios to obtain an estimate of the most likely sequence of events. If individuals are risk averse, the standard principles of expected utility theory (see Kreps, 1990) sugge,st that this approach may provide a poor guide to policy given the substantial uncertainties associated with climate change. One approach to cost-benefit analysis under uncertainty is to use ad hoc procedures to adjust expected outcomes for risk. Since individuals demand relatively high expected rates of return on risky investments, cost-benefit analysts sometimes apply high discount rates in evaluating uncertain projects. Although tractable in practice, this approach raises theoretical objections (Wilson, 1982). A rational investor will demand a high expected return on an uncertain investment if its returns are positively correlated with the return on her/his overall investment portfolio. Conversely, she/he will accept comparatively low (or even negative) expected returns on assets that reduce risk by yielding high returns when the market as a whole turns sour. In principle, cost-benefit techniques may be used to identify social willingness to pay to reduce climatic risks (Manne and Richels, 1991). Suppose that there are i = O,..., n(r) possible outcomes or “states of nature” sZi at dates t = O,..., T. The probability of each state is Pr(s,,), and the costs and benefits of greenhouse gas emissions abatement are C,i and Bti. Then an efficient level of abatement may be identified by maximizing the net present-value expression T
n(t)
NPV = C C Pr( ~,i) S,i( B,i - C,i) t=O i=l
(4)
where the discount factor & varies across time and states of nature., reflecting individuals’ risk aversion and their relative well-being at sequential dates (Howarth and Norg,aard, in press). Exhaustive infonmation would be required to rigorously evaluate this criterion, which flows logically from the definition of intertemporal efficiency under uncertainty. The analyst would need to identify the complete set of possible future states, including their statistical probability, climatic conditions, and implications for human welfare. She would also need to gauge social preferences regarding time and risk, even for low-probability, high-consequence outcomes about which there is little objective information.
and Planetary Change I I (1996) 187-199
195
A heuristic version of this approach is offered by Cline (1992b), who examines the net benefits of greenhouse gas emissions abatement under three scenarios regarding the relationship between greenhouse gas stocks and climate damages: low, medium, and high. Each scenario is assigned a subjective weight based on its probability and the presumed risk aversion of the decision maker. Though such calculations are illustratively useful, they do not meet the demanding information requirements necessary to identify efficient resource allocations in the context of uncertainty. In this sense, cost-benefit analysis and the efficiency criterion are not fully operational in the analysis of climate policies. 6.5. The sustainability criterion Given the limitations of the efficiency criterion in addressing issues of uncertainty and intergenerational fairness, one might posit a social welfare function as a means of comparing and evaluating alternative strategies based on the comparative welfare of present and future generations across the complete range of possible outcomes. In principle, a social welfare function would simultaneously cope with questions of fairness and uncertainty by reducing social values to a single, well-defined criterion. In actuality, efforts to define an appropriate welfare function have failed for well-known theoretical and practical reasons (Sen, 1970). Even if a welfare function were fully identified, using it to compare alternative policies would run up against the same information requirements that confound cost-benefit analysis under uncertainty. We are left then to identify criteria that capture prevailing principles of intergenerational justice under uncertainty. The notion of intergenerational fairness as it is usually put forth in public debates over environmental policy takes the form of a constraint on the range of outcomes that are considered ethically permissible rather than a utilitarian definition of optimality (Howarth, in press). One widely accepted view is that economic and social development should be sustainable in the sense that it “meets the needs of the present without compromising the ability of future generations to meet their own needs” (WCED, 1987, p. 431, ensuring future generations the natural resources and environmental quality re-
196
R.B. Howarth, PA. Monahan/Global
quired to enjoy a favorable way of life. This approach is rooted in the moral premise that members of future generations should have life opportunities no worse than those of the present (Partridge, 1981; MacLean and Brown, 1983; Brown Weiss, 1989). Economists have interpreted sustainability as the technical requirement that the welfare of successive generations should be no lower than that of their predecessors (Pezzey, 1989). Because of the difficulties involved in defining the welfare of future generations, this approach is not directly operational. An alternative is to ensure that per capita consumption is nondecreasing over time. But aggregate economic indicators generally neglect non-market environmental amenities and the degradation and depletion of natural resource stocks (Repetto et al., 1989). Application of this approach would at a minimum require a careful reconsideration of conventional accounting techniques, and would likely run up against steep information requirements (Daly and Cobb, 1989). An alternative approach focuses on the conditions required to support a high standard of living into the indefinite future rather than the distribution of welfare across generations per se. Thus sustainability implies that we should ensure “the ability of future generations to meet their own needs” (WCED, 1987) or that future generations should inherit “an undiminished natural and economic endowment” (Brown et al., 1990). This approach does not require an exact definition of the welfare of future generations. It does, however, imply an obligation to conserve environmental quality for the benefit of future persons. Uncertainty concerning the future course of development constitutes a major concern of the sustainability criterion. How far should we go to protect future generations against the possibility of an inhospitable world? The composition of future generations will depend on the state of the world prevailing when they are born (Parfit, 1983). The individuals alive at a particular date under alternative contingent states should thus be regarded as ethically distinct potential generations, and sustainability would seem to require that the welfare of each potential generation be equal to or greater than that of its predecessor. Thus, in a world of uncertainty, the sustainability criterion may require sacrifices on the part of the present generation not only to raise the expected welfare of future generations but also to ensure that
and Planetary
Change
II (1996)
187-199
living standards are non-decreasing even under pessimistic circumstances. This is a strong supposition that needs to be placed in the context of competing social values. Few would argue, for example, that fifty percent of world income should be diverted to the construction of a planetary defense system to protect against the risk that future generations might be devastated by a collision between the Earth and a large asteroid. On the other hand, the world community has decided to incur significant costs to reduce the uncertain threat posed by ozone depletion in the upper atmosphere. Brown et al. (1990) argue that “A sustainable society is one that satisfies its needs without jeopardizing the prospects of future generations”. At a minimum, this interpretation suggests an obligation to reduce threats to future generations if so doing does not noticeably impact the subjective welfare of existing persons. 7. From normative criteria to climate policy Climate policy formulation involves balancing the interests of present and future generations and peoples of different world regions. A perceived benefit of climate stabilization is the reduction of risks to future generations, particularly those future persons most vulnerable to climate fluctuations because their poverty renders them susceptible to the vagaries of nature. Some analysts hold that the future impacts of climate change are too uncertain to justify the expenditure of present-day resources to mitigate greenhouse gas emissions. Gray and Rivkin (1991) argue for a “no-regrets” policy, whereby mitigation measures are implemented only if they yield clear benefits in ameliorating other social problems. A no regrets strategy would embrace the phase-out of CFCs under the Montreal Protocol given the chemicals’ role in destroying stratospheric ozone. Tropical deforestation is caused in part by policies that promote land-use changes through direct and indirect subsidies. Similarly, many nations subsidize the use of fossil fuels. A no-regrets strategy would reduce such subsidies to improve economic efficiency while at the same time slowing climate change. It would also embrace measures facilitating the adoption of cost-effective energy-efficient technologies.
R.B. Howarrh, PA Monahan/Global
The achievement of substantial long-term reductions in greenhouse gas emissions, however, implies moving beyond no-regrets measures to policies with costs ranging from ambiguous to rather large. Stabilization of today’s climate would require immediate carbon dioxide emissions reductions of some 60% relative to today’s level. Such reductions are arguably politically and socially infeasible. The question is thus not whether we should permit climate change or not, but what steps (if any> should be taken today to reduce the magnitude and timing of change. Krause et al. (19139)argue that human and ecological systems could adapt to temperatures 2.5”C above the pre-industrial norm if the rate of change was limited to O.l”C per decade. Based on this supposition, they identify maximum permissible greenhouse gas concentrations for each future date and work backwards to identify emissions constraints. They conclude that global emissions of carbon dioxide in industrialized nations should be reduced by 20% through 2005 with reductions of 50% and 75% achieved by 2015 and 2030. To support the sustained improvement of living standards, developing countries would be permitted to increase emissions by 50-100% over the short term, with emissions retuming to current levels by 2030. This approach is generally consistent with the United Nations Framework Convention on Climate Change, which calls for (Article 2): “[the] stabilization Iof greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the climate system. Such a level should be achieved within a time frame sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not threatened and to enable economic development to proceed in a sustainable manner”. The agreement mandates industrialized nations to return ‘‘anthropoge nit emissions of carbon dioxide and other greenhouse gases not controlled by the Montreal Protocol” to their 1990 levels by the year 2000 (Article 4, paragraph 2(b)). Reducing carbon emissions in the United States by 20% by the year 2010 would require the imposition of a carbon 1:ax of perhaps $100 per tonne (Howarth and Winslow, 1994). Such a tax would likely reduce economic activity by some O-2%, de-
and Planetary Change I1 (1996) 187-199
197
pending on its coordination with other policies 3. One could not credibly advance a carbon tax as a no-regrets approach to climate stabilization. The impacts of moderate emissions abatement on subjective human welfare, however, are somewhat ambiguous. Suppose that a 20% reduction in carbon dioxide emissions by the year 2010 would reduce economic activity by 1% in the industrialized world. Phased in gradually over time, this change would reduce economic growth by about O.O7%/yr. Except for workers in directly affected sectors such as coal mining and the supply of energy-efficient equipment, these changes might well pass unnoticed by most members of society. Perceived losses, where they occurred, could presumably be compensated through job retraining and other programs appropriate in a dynamic economy characterized by structural change. We have established that greenhouse gas emissions, left unmitigated, threaten uncertain but potentially serious impacts on the welfare of future generations. Catastrophic change is a distinct possibility that is consistent with prevailing knowledge of the interactions between climate and biogeochemical systems. There is good reason to believe that the costs of reducing greenhouse gas emissions in the industrialized nations are today zero or negative at the margin, while moderate abatement would impose modest costs with limited impacts on the subjective well-being of most individuals. To move from these premises to clear policy recommendations requires use of a corollary of the sustainability principle: Inhabitants of today’s world are morally obligated to take steps to reduce catastrophic risks to members of future generations if doing so would not noticeably diminish their own quality of life. This form of argument does not require a precise characterization of the impacts of climate change on future society, nor does it imply that the probabilities of extreme events must be calculated with confidence. Instead, we construct a simple two-part test based on the concept of operational thresholds. Is there a non-trivial risk of catastrophic change? Can
3Zero net abatement costs might possibly arise given optimistic assumptions concerning the use of carbon tax revenues to offset distortionaty taxes and the adoption of cost-effective energy-efflcient technologies. See section IV above.
198
R.B. Howarth, PA. Monahan /Global and Planetary Change II (1996) 187-199
we reduce that risk without unduly compromising our own well-being? The greenhouse gas emissions guidelines embodied in the Framework Convention would seem to pass both parts of this test. The sustainability principle also sheds light on the appropriate sharing of burdens between North and South. While the primacy of the sustainability rule is sometimes attacked on the grounds that we should take care of today’s poor before worrying about the future, principles of intergenerational justice flow logically from principles of justice between contemporaries (Malnes, 1990; Howarth, 1992)-the distinction between the two is based on a false premise. The Framework Convention holds that the costs of climate stabilization should be borne by those most able to manage them-the wealthy nations of the industrialized North. Indeed, climate stabilization arguably implies the transfer of technology and other assets from North to South to permit sustained development while limiting the growth of greenhouse gas emissions. A number of questions remain for further research. To argue that the sustainability criterion supports immediate steps towards greenhouse gas emissions abatement leaves open the question of the extent of emissions reduction that might be required over the longer term. The practical role of cost-benefit analysis in supporting climate policy is also worthy of further consideration. In principle, intertemporal efficiency and intergenerational fairness are complementary, not contradictory, criteria for resource allocation. In practice, economic analysis provides a means of modeling the impacts of policy measures on indicators such as employment and national income growth while offering concrete guidance as to how social objectives concerning environmental quality can be achieved at minimum social cost. References Broecker, W.S., 1987. Unpleasant surprises in the greenhouse. Nature, 328: 123- 126. Broome, J., 1992. Counting the Cost of Global Warming. White Horse Press, Cambridge. Brown, L.R., Flavin, C. and Postel, S., 1990. Picturing a sustainable society. In: Worldwatch Institute, State of the World 1990. Norton, New York, pp. 173-190.
Brown, P.G., 1992. Climate change and the planetary trust. Energy Policy, 20: 208-222. Brown Weiss, E., 1989. In Fairness to Future Generations: International Law, Common Patrimony, and Intergenerational Equity. Transnational, Dobbs Ferry, New York. Cline, W., 1992a. Global Warming: The Economic Stakes. Inst. Int. Econ., Washington. Cline, W., 1992b. The Economics of Global Warming. Inst. Int. Econ., Washington. Daly, H.E. and Cobb, J.B., 1989. For the Common Good: Redirecting the Economy Toward Community, the Environment, and a Sustainable Future. Beacon Press, Boston. Dansgaard, W., White, J.W.C. and Johnsen, S.J., 1989. The abrupt termination of the Younger Dryas climate event. Nature, 339: 532-534. Gleick, P.H., 1992. Effects of climate change on shared fresh water resources. In: I. Mintzer (Editor), Confronting Climate Change: Risks, Implications and Responses. Cambridge Univ. Press, pp. 127- 140. Gray, C.B. and Rivkin Jr., D.B., 1991. A “no regrets” environmental policy. Foreign Policy, N83: 47-64. Howarth, R.B., 1992. Intergenerational justice and the chain of obligation. Environ. Values, 1: 133- 140. Howarth, R.B., in press. Sustainability under uncertainty: a Kantian approach. Land Economics. Howarth, R.B. and Monahan, P.A., 1992. Economics, Ethics, and Climate Policy. Stockholm Environ. Inst., Stockholm. Howarth, R.B. and Norgaard, R.B., 1992. Environmental valuation under sustainable development. Am. Econ. Rev., 82: 473-477. Howarth, R.B. and Norgaard, R.B., in press. Intergenerational choices under global environmental change. In: D. Bromley (Editor), Handbook of Environmental Economics. Blackwell, Oxford. Howarth, R.B. and Winslow, M.A., 1994. Energy use and CO, emissions reduction: integrating pricing and regulatory policies. Energy, 19: 855-867. Hulm, P., 1989. A climate of crisis: global warming and the South Pacific islands. Assoc. South Pac. Environ. Inst., Port Moresby, Papua New Guinea. IPPC (Intergovernmental Panel on Climate Change), 1991a. Climate Change: The IPCC Response Strategies. Island Press, Washington. IPCC (Intergovernmental Panel on Climate Change), 1991b. Climate Change: The IPCC Scientific Assessment. Cambridge Univ. Press. Kolb, J.A. and Scheraga, J.D., 1990. Discounting the benefits and costs of environmental regulations. J. Policy Analys. Manage., 9: 381-390. Krause, F., Bach, W. and Koomey, J., 1989. Energy Policy in the Greenhouse, Volume I. Int. Project Sustainable Energy Paths, El Cenito, CA. Kreps, D.M., 1990. A Course in Microeconomic Theory. Princeton Univ. Press. Levy, F. and Mumane, R.J., 1992. U.S. earnings levels and earnings inequality: a review of recent trends and proposed explanations. J. Econ. Literature, 30: 1333-1381.
R.B. Howarth, PA. Monahan/Global Lewis, J., 1989. Sea level rise: some implications for Tuvalu. Ambio, 18: 458-459. Lind, R., 1982. Discounting for Time and Risk in Energy Policy. Resour. Future. MacLean, D. and Brown, PG., 1983. Energy and the Future. Rowman and Littlefield, Totawa, NJ. Malnes, R., 1990. The Environment and Duties to Future Generations: An Elaboration of Sustainable Development. Fridtjof Nansen Inst., Lysaker. Manne, A.S. and RicheH, R.G., 1991. Buying greenhouse insurance. Energy Policy, 6: 543-552. Mearns, L.O., Katz, R.W. and Schneider, S., 1984. Extreme high temperature events: changes in their probabilities with changes in mean temperatures. J. Clim. Appl. Meteorol., 23: 16011613. Nordhaus, W.D., 1989. The economics of the greenhouse effect. Paper presented to the MIT Workshop on Energy and Environmental Modeling and Policy Analysis. Nordhaus, W.D., 1990. Economic policy in the face of global warming. Dep. Econ., Yale Univ. Nordhaus, W.D., 1991a. Economic approaches to global warming. In: R. Dombusch and J. Poterba (Editors), Global Warming: Economic Policy Responses. MIT Press, Cambridge, MA. Nordhaus, W.D., 1991b. To slow or not to slow: the economics of the greenhouse effect. Econ. J., 101: 920-937. Nordhaus, W.D., 1992. An optimal transition path for controlling greenhouse gases. Science, 258: 13 15- 13 19. Nordhaus, W.D., 1993. Reflections on the economics of climate change. J. Econ. Perspcct., 7: 1 l-25. Oeschger, H. and Mintzer, I., 1992. Lessons from the ice cores: rapid climate changes during the last 160,000 years. In: I. Mintzer (Editor), Confronting Climate Change: Risks, Implications and Responses. Cambridge Univ. Press, pp. 55-64. Parfit, D., 1983. Energy policy and the further future: the identity problem. In: D. MacLean and P.G. Brown (Editors), Energy and the Future. Rowman and Littlefield, Totawa, NJ, pp. 38-58. Parry, M.L., 1990. Climate Change and World Agriculture. Earthscan, London. Partridge, E., 1981. Responsibilities to Future Generations. Prometheus, Buffalo. Pezzey, J., 1989. Economic Analysis of Sustainable Growth and Sustainable Development. World Bank, Washington. Randall, A., 1987. Resource Economics: An Economic Approach to Natural Resource and Environmental Policy. Wiley, New York, 2nd ed. Repetto, R., Magrath, PI., Wells, M., Beer, C. and Rossini, F., 1989. Wasting Assets: Natural Resources in the National Income Accounts. World Resour. Inst., Washington.
and Planetary Change II (1996) 187-199
199
Revelle, R., 1983. Methane hydrates in continental slope sediments and increasing atmospheric carbon dioxide. In: National Research Council, Changing Climate. Natl. Acad. Press, Washington, pp. 252-261. Rind, D., Goldberg, R., Hansen, J., Rosenzweig, C. and Ruedy, R., 1990. Potential evapotranspiration and the likelihood of future drought. J. Geophys. Res., 95: 9983-10,004. Rose, A. and Stevens, B., 1993. The efficiency and equity of marketable permits for CO, emissions. Resource Energy Econ., 15: 117-146. Schelling, T.C., 1992. Some economics of global warming. Am. Econ. Rev., 82: 1-14. Schimel, D., 1990. Biogeochemical feedbacks in the earth system, In: J. Leggett (Editor), Global Warming: The Greenpeace Report. Oxford Univ. Press. Schneider, S., 1989. Global Warming. Sierra Club Books, San Francisco. Sen, A.K., 1970. Collective Choice and Social Welfare. HoldenDay, San Francisco. Shackleton, R., Shelby, M., Cristofaro, A., Brinner, R., Yanchar, J., Goulder, L., Jorgenson, D., Wilcoxen, P. and Pauly, P., 1992. The efficiency value of carbon tax revenues. U.S. Environ. Protect. Agency, Washington. Svedin, U. and Aniansson, B., 1987. Surprising Futures. Swedish Count. Plann. Coord. Res., Stockholm. USEPA (U.S. Environmental Protection Agency), 1989a. Policy Options for Stabilizing Global Climate. Government Printing Office, Washington. USEPA (U.S. Environmental Protection Agency), 1989b. The Potential Effects of Global Climate Change on the United States. Government Printing Oftice, Washington. USEPA (U.S. Environmental Protection Agency), 1990. The Potential Effects of Global Climate Change on the United States. Hemisphere, Washington. U.S. National Academy of Sciences, 1991. Policy Implications of Greenhouse Warming: Report of the Mitigation Panel. Natl. Acad. Press, Washington. Weyant, J.P., 1993. Costs of reducing global carbon emissions. J. Econ. Perspect., 7: 27-46. Wilson, R., 1982. Risk measurement of public projects. In: R. Lind (Editor), Discounting for Time and Risk in Energy Policy. Resour. Future, pp. 205-249. Woodwell, G.M., 1990. The effects of global warming. In: J. Leggett (Editor), Global Warming: The Greenpeace Report. Oxford Univ. Press, New York, pp. 116- 132. WCED (World Commission on Environment and Development), 1987. Our Common Future. Oxford Univ. Press, New York.