One of the most critical issues in policy design is comparing and assessing different trajectories to achieve GHG emissions reductions and evaluating the consequences and implications of those trajectories for human and environmental systems. A recent NRC study (NRC, 2010j) examined the implications for a range of climate stabilization targets. In contrast, this subsection provides a high-level overview of the social science research needs associated with analytic methods to evaluate targets, focusing on the two major alternative approaches: benefit-cost analysis and cost-effectiveness analysis.
Benefit-cost analysis is a method of systematic evaluation of the total social consequences of any decision or strategy. Applied to climate change, it has been used to assess alternative GHG emissions trajectories, typically by comparing a few simple al ternative trajectories—each with associated projections of GHG concentrations, global average temperature, climate change impacts, and their valuation—with projections of the cost and effort required to achieve the trajectory relative to some baseline. Benefit-cost analysis requires expressing climate change impacts and lost services in an overall monetary metric so they can be compared to estimates of the costs associated with policies to limit the magnitude of future climate change. Ancillary costs and co-benefits of climate policies, which refer to costs and benefits in other areas (such as changes in local air pollution) resulting from climate policies, are sometimes included in the calculations as well. Such analysis can be conducted for a specific region or nation, or for the world.
Benefit-cost analysis can examine the expected benefits and costs of a particular target or policy or, by looking across targets, can identify an optimum target that maximizes net social benefits. If costs and benefits can be systematically and reliably projected and compared (discussed below are five of the major challenges that must be met to accomplish this objective), the socially optimal level of GHG emissions will be the level where the marginal benefit of reducing GHG emissions further will be equal to the marginal cost of making further GHG emissions reductions. If these calculations can be done credibly, decision makers can use this information to (1) set a limit on GHG emissions, (2) set a price on GHG emissions (whether implemented through market mechanisms or full costing of regulatory programs like emissions standards), and (3) get some sense of how important the climate problem is relative to other major societal problems.
An alternative approach, cost-effectiveness analysis, stipulates some limit on climate change (e.g., a future limit on human-caused radiative forcing or global-average temperature change) as a fixed goal, without evaluating the climate damages associated with the goal, then compares the costs of alternative emissions and policy trajectories to achieve that goal. For example, cost-effectiveness analysis has been used to compare alternative trajectories by which global emissions slow their growth and then decline to meet specified limits on atmospheric CO2 concentration or human-caused radiative forcing in the 22nd century (e.g., starting the decline immediately versus growing for a decade or two and then declining faster [CCSP, 2007c; Richels et al., 2007; van Vuuren et al., 2006; Wigley et al., 1996]). A variant of cost-effectiveness analysis that has been used for climate change, called "safe landing" or "tolerable windows" analysis, defines two such constraints, one on the amount (and sometimes the rate) of climate change and another on the maximum rate of global emissions reduction. It then examines the cost and feasibility of alternative trajectories that stay within those boundaries (Fussel et al., 2003).
Cost-effectiveness approaches are often used when the costs and benefits of some action differ greatly in character, and the benefits are subject to greater uncertainty or controversy. In this circumstance, cost-effectiveness analysis allows analytically based comparisons of decisions without requiring that all impacts—in this case, damages from climate change and costs of emissions reduction—be reduced to a single metric. However, the implicit value this imputes to GHG emissions reductions is still equal to the marginal cost of GHG emission reductions that results from hitting the target or staying within the tolerable window. Of course, this implicit value can then be used to adjust the target if that value is felt to be lower or higher than the aggregated marginal value of the climate change impacts avoided. Such an iterative approach to GHG target setting allows multiple metrics to be used in evaluating the impacts of climate changes without completely abandoning the discipline provided by the strict application of cost-benefit analysis.
Formal policy-analytic methods such as benefit-cost and cost-effectiveness analysis can be powerful tools for informing decisions and illuminating structural issues underlying them and have had significant influence in climate policy debates. However, in practice, several major challenges must be met to provide reliable guidance to policy (e.g., Adler and Posner, 2006; Atkinson and Mourato, 2008; Dietz, 1994; Graves, 2007). The modeling community has made significant progress in addressing each of these challenges (see references within each section), but further progress in each area would greatly improve the usefulness of the results produced.
Five challenges are particularly difficult and influential in contributing to differences in cost-benefit valuations between alternative studies. These challenges, discussed in the paragraphs below, have to do with being able to systematically and comprehensively evaluate the benefits of GHG emissions reductions, being able to consistently and comprehensively project the costs of GHG emissions reductions, or being able to compare costs and benefits over time, under uncertainty, and across different socioeconomic groups.
1. Estimating the social value of goods and services, particularly for impacts on ecosystems, climate-related amenities, or other resources and values for which market prices do not exist. If formal policy-analytic methods are to be used to inform the choice of climate targets, rather than merely the choice of alternative means to meet a specified target, then all consequences of climate change and of efforts to limit it must be made comparable and valued. Economic theory argues that prices in well-functioning markets reflect the full social value of the goods and services that are exchanged, so market prices can be used to value changes in those goods caused by climate change. For impact sectors where markets exist like agriculture, this allows structural models calibrated with market data to be used to value changes in activity levels that result from climate change. However, many important things that will be affected by climate change, such as environmental amenities, ecosystem services, and human health effects, are not exchanged in markets and so have no market price to provide guidance on their social valuation. This problem is pervasive in many areas of environmental assessment, and various methods, including contingent valuation and hedonic pricing approaches, have been developed to infer people's valuations for nonmarketed goods from their choices in related markets or suitably disciplined surveys (Arrow et al., 1993; Atkinson and Mourato, 2008; Carson, 1997; Mendelsohn and Olmstead, 2009). However, the range of estimates from the various studies is large and there is no consensus on the best approaches.
2. Valuing uncertain outcomes, particularly high-consequence events whose probability is believed (but not known) to be low at low levels of warming, but increases with greater climate forcing, often called the problem of "fat tails." Outcomes like these could plausibly result from dramatic irreversibilities in the climate or climate-impacted systems (e.g., a large ice sheet like Greenland melts very rapidly, increasing sea levels and reducing the reflection of sunlight from it, or large amounts of GHGs are released from warming permafrost). In principle, uncertain outcomes can be given a probability weight so that more likely outcomes are given greater weight and less likely—but much worse—outcomes are given lesser weight. "Fat tails" then generally refers to the case where the probability of very-high-consequence outcomes is still high enough that the product of that probability times the valuation of climate damages that would result from that outcome is large (i.e., does not approach zero because the probability of the outcome goes to zero more slowly than the impact valuation of that outcome increases). At present, it is difficult to estimate the probabilities of uncertain climate outcomes, but when these uncertainties are included in an analysis, the results can be very sensitive to assumptions that are made about the probability distribution associated with these low-probability/high-consequence events, and result in quite different conclusions (Nordhaus, 2009; Stern, 2007; Weitzman, 2007b, 2009; Yohe and Tol, 2007). Equally or more important here can be assessing how people perceive and act on the different risks that they face.
3. Comparing costs, damages, and impacts in the near and long term (setting a social discount rate). The rationale for discounting future costs and benefits (i.e., assigning them a lower value than immediate ones) has been discussed for more than 80 years (see Portney and Weyant  for an overview). Discounting has both an ethical and a scientific component, and when these are correctly distinguished, the case for some form of discounting is compelling (Arrow et al., 2004; Heal, 1997; Nordhaus, 2008; Weitzman, 2007a). There are substantial disagreements, however, over the appropriate functional form and quantitative magnitude of discount factors, and whether it is appropriate to apply the same approach to monetary costs and environmental damages (Atkinson and Mourato, 2008; Dasgupta and Ramsey, 2008; Dietz and Stern, 2008; Graves, 2007; Heal, 2009; Yohe, 2006). In addition, since many of the people who will be affected by climate change the most have not yet been born, an equitable way of factoring their preferences, which cannot be directly measured, into the calculations needs to be developed (Portney and Weyant, 1999).
Because many costs of reducing climate change occur in the near term while the most serious of the climate impacts avoided would be further in the future, socially optimal levels of climate change limitation in a benefit-cost framework can be quite sensitive to choices about discounting; lower discount rates usually imply stronger and earlier action to limit climate change than higher discount rates.
4. Estimating how policy will influence technological change. It is possible that technological innovation will create opportunities to reduce GHG emissions at lower than present costs, but the rate of such innovation and the relative influence and mechanisms of various possible ways to stimulate it are subject to substantial uncertainties. Alternative models of induced technological change highlight the influence of policies to raise the effective price of emissions, learning-by-doing, public versus private investments in research and development structured in various ways, basic science versus specifically targeted research, and overall investment driven by aggregate economic growth. These alternative models can imply substantial differences in preferred policies, but available data appear to not discriminate strongly between them (Goul-der, 2004; Grübler, et al., 2002).
5. Incorporating equity considerations into the analysis. The costs and benefits of climate change adaptation and limitation will be unevenly distributed across space, time, and social and economic groups. There will be substantial differences across regions within the United States and across the globe (USGCRP, 2009a; World Bank, 2009). Although costs and benefits could in principle be weighted to incorporate equity concerns (Atkinson and Mourato, 2008; Kverndokk and Rose, 2008), in practice this poses significant challenges of observing and projecting disaggregated costs and benefits and, if aggregation is required, identifying defensible equity-based weights. Moreover, formal analyses of climate change responses have examined only aggregate effects at the level of the jurisdiction considered. International aggregations of climate change impacts are often valued in terms of losses in income, which tends to bias the weighting toward richer and away from poorer people who have less to lose but will feel percentage losses in income more.
That there are significant research challenges that remain regarding the major ele ments of cost-benefit analysis as currently applied to climate change policy evaluation means that great care needs to be exercised in communicating the results of these calculations to decision makers. Consumers of these studies need to know what factors are included in the analysis and how, and which ones are left out or only partially represented. They can then add their own assessments of the missing elements and perspectives to the numbers they get from the cost-benefit calculations in order to provide a more complete picture of the value of different policy alternatives. At the same time, the raw numbers themselves, if interpreted correctly, can often help decision makers set bounds on appropriate actions, especially if we are far away from the optimum.
Was this article helpful?