The Uses and Misuses of Technology Development as a Component of Climate Policy
The current misplaced focus on short-term climate policies is a product both of domestic political exigencies and badly flawed technical analyses. A prime example of the latter is a recent U.S. Department of Energy study, prepared by five national laboratories in late 1997. The Department’s “Five-Labs” study assumes-incorrectly-that technical solutions are readily at hand. Also, it wrongly suggests that a short-term technology fix is easily available at little or no cost. Worse, advocates of short-term emissions targets under the Framework Convention on Climate Change are using this study to justify the subsidy of existing energy technologies-diverting resources from the effective long-term technology response that will be needed if the climate picture darkens.
The search for cleaner energy technologies is central to any long-term response to the threat of global climate change. Unfortunately, most climate debates focus on short-term Kyoto Protocol-type emissions targets and Congressional reaction to that approach. This is a tragic mistake. Valuable time is being lost that could better be spent searching for cost-competitive, low-carbon energy sources that may be needed in the future.
The misplaced focus on short-term policies is a product both of domestic political exigencies and badly flawed technical analyses. A prime example of the latter is a recent study of technology options prepared for the Administration by five U.S. Department of Energy national laboratories. The Department’s “Five-Labs” study argues-incorrectly-that technical solutions are readily at hand that can, at no cost, yield major reductions in U.S. greenhouse gas emissions.
Taken seriously, as it has been, this study contributes to a mis-direction of the technology component of climate policy, and the consequences for the future could be serious. At present, the risk of human influence on climate appears substantial. No doubt there is great uncertainty about the possible magnitude of potential change. Our own Massachusetts Institute of Technology analysis shows that the range of possible outcomes over the next century is very wide.1 We have only a partial understanding of the behavior of the climate system, and our ability to forecast the future path of economic growth and technical change, and the resulting greenhouse emissions, is naturally limited. Our understanding of the ecosystem effects of rapid climate change is even weaker. But this type of uncertainty means that the problem may turn out to be either not so important, or worse than we imagine. Thus, whatever one’s position regarding the current conflict over near-term emissions controls, we should be able to agree that we are taking considerable risks with global climate in the long term. Further, we should anticipate that increased knowledge could well lead to general international agreement that stringent measures are needed.
Unfortunately, even universal consensus would not necessarily lead to meaningful action. Parties to any treaty will have very different national priorities and interests, and it is difficult to imagine an agreement that could suppress the incentive to be a free rider on the efforts of others. Also, while rich countries could (in principle) use various forms of foreign aid, emissions permit trading, or other side payments to encourage other countries to participate, the amount of money involved would be very large, both by historical standards and in the eyes of most politicians and taxpayers.
Thus it may be that the key to future response capability is the development of low-carbon technologies capable of competing head-to-head with conventional fossil-fuel burning, or technologies that require only small cost penalties. These technologies are not now available, and we cannot expect resource depletion to give them a cost advantage over conventional fossil fuels. With limited R&D resources, and with political attention focused on short-term policy targets, the United States runs the very real risk of forfeiting promising long-term technology development-perhaps to our considerable regret later. Understanding the flaws in Five-Labs­p;type studies is a key to getting the climate debate back on track.
Two Approaches to Analyzing Emissions Control Policies
Two approaches commonly are used to analyze emissions control policies: market-based analysis and technology-costing. The market-based approach is favored by economists because it focuses on market price as the essential mechanism through which policy affects economic activity. This “top-down” analysis examines the entire economy (or a number of interacting sectors) and the interplay of specific policies within the larger economic system. By contrast, the technology-costing or “bottom-up” approach tries to estimate the cost of particular technologies and/or devices, totals the energy savings (or emissions reductions) from these technologies, and compares such costs to reduced energy expenditures and/or other environmental benefits.
Both analytic approaches have strengths and weaknesses. Market-based methods capture input costs (such as capital, labor, and natural resources) and reflect, at least to some degree, consumer preferences and how such preferences change as wealth increases. There is a trade-off, of course. By relying on simplified representations of production processes and aggregation of industrial sectors, market-based analyses sacrifice some technical detail. On the other hand, market interactions are crucial to capturing the size of economic change that would result from climate policies now being considered. Thus, so long as its unavoidable uncertainty is appropriately expressed, a market-based approach is superior to a technology-costing analysis of the Kyoto Protocol.
Pitfalls of Technology-Costing
Technology-costing studies often suffer from one or more of the following shortcomings:
Confusing Market Failures With Market Barriers
Technology-costing studies often begin by listing individual technologies or devices that save energy or otherwise look attractive but are “under-used,” either because of “market failure” or “market barriers.”2 A market failure occurs when some flaw in the way markets are organized causes consumers and/or producers to respond to the wrong price signals. Such failures include situations where decision makers have insufficient or incorrect information (consumers do not perceive how costly it is to own an power-guzzling air conditioner), where there is an asymmetry between the person making the energy-use decision and the one paying the bills (the landlord-tenant problem), or where there is some form of non-priced externality (such as urban air pollution).
Market barriers, often resulting from buyer preferences, also may retard the spread of a technology that has positive net present value (NPV) on an engineering cost basis. Several reasons may lie behind this behavior. Potential buyers simply may not like the technology. Or there may be transactions costs or other expenses of adoption that are hard to include in the calculations. Uncertainty about performance, too, or about energy prices, may lead users (quite rationally) to exercise their option to wait and reconsider the technology next year. And, importantly, in their day-to-day decisions buyers may use discount rates higher than those assumed by the cost analysts. The key difference between market barriers and market failures is that correcting failures may sometimes produce a net benefit, whereas overcoming barriers always involves cost.
Lack of Attention to Market Structure
Incomplete analysis-not consumer behavior-is another reason why analysts’ expectations so often exceed actual market penetration of a technology. After all, it is not easy to assess factors other than engineering NPV that determine whether-and how fast-a new technology is adopted by consumers and businesses. These factors include the structure of distribution channels, other industries that supply key (and perhaps new) inputs, regulatory and legal-liability issues (such as health, safety, and environmental quality), and, most important, the internal organization of the industry at which the new technology is aimed.
Failure to Account for Inter-Market Adjustments
Finally, analysis that looks at individual technologies and the policies that might aid their entry into the marketplace often overlooks the many complex interactions across markets. If these changes are small, market interaction can be ignored. But if changes are large in relation to the markets at issue, then the analysis can be deeply flawed. Proposed climate policies, such as the Kyoto Protocol, clearly involve major shifts in the structure of energy supply and use. An obvious source of concern, therefore, is how analysts evaluate large-scale policy-induced changes in the use of a particular fuel or other input, especially when a study assigns an exogenous (and constant) estimate to domestic energy prices.
The Department of Energy Five-Labs Study
A late-1997 analysis of U.S. carbon reductions prepared by a consortium of five national government laboratories3 begins with an observation that many energy-efficient technologies remain “underutilized,” then sets out to “quantify the reductions in carbon emissions that can be attained through the improved performance and increased penetration of efficient and low-carbon technologies by the year 2010.” As stated in the report, if the policy measures discussed are taken, “the cases analyzed in the study are judged to yield energy savings that are roughly equal to or greater than the costs.” How the authors arrived at this conclusion is worth examining since the faulty assumptions and logic behind their work threaten to undermine the very foundations of energy R&D efforts needed to foster effective long-term climate technology development.
The Study’s Structure
The Five-Labs study divides the U.S. energy economy into four sectors (buildings, industry, transportation, and electric utilities), and estimates how much a technology push might reduce carbon emissions in 2010. Reductions in energy and carbon emissions are calculated from a baseline or “business as usual” case drawn from forecasts in the Energy Information Administration’s 1997 Annual Energy Outlook. Possible carbon reductions are compiled for three policy cases:
- Efficiency Case. This case presumes “an invigorated public- and private-sector effort to promote energy efficiency through enhanced R&D and market transformation activities.”
- High-Efficiency/Low-Carbon (HE/LC) Case with a $25 price per ton of carbon ($/tC). This case presumes that the carbon premium is attained through a cap-and-trade system. This policy is accompanied by a “greater commitment” through federal programs, strengthened state programs, and “very active” private sector involvement.
- High-Efficiency/Low-Carbon (HE/LC) Case with a $50/tC price. This case doubles the carbon price but is otherwise the same.
Under the Department of Energy’s pre-Kyoto “business as usual” forecast, returning to 1990 emissions levels would require a decline of about 390 million tons of carbon (MtC) annually.4
The Study’s Shortcomings
Two general points should be made about the analytic method used in the Five-Labs study. First, only in the electric utility sector is the potential impact of carbon prices actually quantified. For all other sectors, the net decreases claimed are the result of judgments about the effects of carbon price, supposedly augmented by a “greater commitment” by government and a “very active” private sector. Second, the Five-Labs study does not describe the policies that might be used to increase market penetration of energy-efficient or low-carbon technologies.
Then there are the details of the calculations for individual sectors. A closer look at just three of the many components of this study-automobile fuel efficiency, building efficiency, and the reduced use of coal for electric power generation-indicates why the analysis cannot support its main conclusion, that a return U.S. carbon emissions in 2010 to 1990 levels could be costless.
Automobile Fuel Efficiency
The Five-Labs study asserts that stiffer fuel efficiency standards will result in substantial reductions in carbon emissions by 2010. Of the 73 MtC reduction in its efficiency case, 61 MtC supposedly results from increased vehicle efficiency and most of the remainder from the introduction of ethanol from biomass (see Table 1). Note that the efficiency increase is the result of the assumed introduction of new technologies, including direct-injection stratified charge engines, direct injection diesels and gasoline- and/or diesel-electric hybrids, as well as changes in vehicle design and materials.
This raises an important question: Why do these technical improvements appear in the Efficiency Case but not in the baseline when no price incentives, such as higher gasoline prices, exist to spur their introduction? The answer comes in two parts. First, vastly increased R&D activity is assumed to reduce by 25 percent the time required for market introduction of new technologies (the mechanism is not explained). And, most important:
[The study] assumes that policies necessary to draw energy-efficiency technology into the market are implemented as needed … [including] fuel economy standards, revenue-neutral feebates, fuel taxes, public information, or some other initiative…”.
That is, underlying the analysis which is supposedly about technology is a regulatory and tax program. In effect, regulation and tax programs are treated as freely available political options-and free of economic costs as well.
Building Efficiency: Arbitrary Assumptions About Markets
The Five-Labs study’s handling of the building sector provides an equally disturbing example of how a failure to examine the structure of markets can also lead to false conclusions. The authors first constructed an estimate of how much energy and carbon would be saved if maximum cost-effective energy improvements were installed in 100 percent of U.S. homes and commercial and public buildings. Improvements assumed include consumer appliances, heating and air conditioning equipment, and building design.
For the Efficiency Case, the Five-Labs study assumes that 35 percent of this maximum reduction is achieved (see Table 1). Why the efficiency case should differ from the baseline in this way (with no price incentives) is not discussed, except to say that the expected savings result from a “moderately vigorous effort” to reduce energy use by a “…combination of mechanisms that may include higher prices…energy-efficiency standards and information programs.” The HE/LC cases assume that 65 percent of the maximum is achieved. Again, no discussion is offered about how energy prices might respond to a carbon permit system. Rather, the 65 percent result flows from a judgment as to the effect of a “vigorous effort” to reduce energy use (in contrast to a “moderately vigorous” effort in the Efficiency Case). That is the extent of Five-Labs analysis.
|Table 1||Forecast for Carbon Reduction Below the 2010 Baseline
(Millions of metric tons of carbon)
|Efficiency||High Efficiency/Low Carbon|
Electricity: Ignore Inter-Market Adjustments
The electric sector is the one area in which the study authors do examine the potential impact of a cap-and-trade program. This sector accounts for one-third of the total reduction needed to achieve the 390 MtC emissions reduction in 2010 (see Table 1). In turn, roughly 70 percent of the 136 MtC electric sector emissions decline results from shifts away from coal generators.
Keeping our eyes on coal switching for a moment, it is worth noting that Table 1 reports calculations from only one emissions permit system, i.e., the one yielding a carbon price of $50/tC. (The $25/tC case was constructed using analysts’ judgment). To arrive at this number, a base case was constructed that approximated the results of the Energy Information Administration’s 1997 Annual Energy Outlook and an attempt was made to adjust the model for market changes that might accompany electricity deregulation. In short, the $50/tC carbon case considers two adjustments that influence coal burning. First, the demand for electric generation is reduced because lower end-use is projected in the building and industry sectors. The lower demand allows reallocation of generation among existing units, away from coal to less carbon-emitting natural gas. Second, whatever the demand level there is assumed to be a re-powering of coal-fired units with natural gas.
At this point, the study’s authors fall prey to the third pitfall described above, namely, they assume that the prices of oil, natural gas, and coal remain constant across all cases. This means that the coal-gas price ratio, which is a crucial input to the calculation of coal-to-gas conversion, also is constant. However, in the real marketplace, if climate policy forces coal demand to fall domestically (electricity generation accounts for about 80 percent of U.S. coal use), coal prices also will fall. Moreover, the conversion of coal plants to natural gas will cause a substantial increase in gas demand, driving up its price. In the Five-Labs analysis, omitting these inter-market effects results in an overestimate of the amount of carbon that would be removed through re-dispatching and re-powering electric power plants. In short, technology-costing analysis is ill suited, as an analytic methodology, to capture important aspects of the policy question it attempts to address.
In its 1999 budget request, submitted in February, 1998, the Administration included a $6.3 billion Climate Change Technology Initiative (CCTI). In keeping with the short-term focus of policy discussions, and Five-Labs-type technological optimism, the CCTI consists principally of tax cuts and other measures aimed at spurring market adoption of existing technologies, not true long-term R&D. Worse, this focus on short-term emissions goals could result in a shrinking of the CCTI itself, including that portion that does contribute to long-term energy R&D. This is a very real risk. Members of the U.S. Congress fear “backdoor implementation” of the Kyoto Protocol, circumventing a Senate resolution that sets forth ground rules for ratification, and R&D is caught up in the conflict.
Given these political realities, a truly long-term energy technology initiative is all the more necessary. Technology-costing studies, such as the Five-Labs study, can provide useful guidance to the process. They can help sort out the most important targets for R&D, identify policies that would advance a technology, and provide some justification for advancing technologies that are almost ready to compete in the marketplace. However, these studies can be misused by analysts and policymakers. For example, the Five-Labs study wrongly suggests that a short-term technology fix is readily available at little or no cost. Science does not run on the policymakers’ clock. Technology takes time to develop and markets require time to respond to new consumer devices. Time also is needed to refine industrial processes and to develop new distribution and materials supply systems. Furthermore, capital turnover varies by sector, particularly in capital-intensive energy supply sectors.
In short, the Five-Labs study misguides the policymaking process by applying a long-range policy tool to short-term emissions goals. Our long-term outlook can be hopeful, provided a well-designed effort is launched and sustained, making adjustments as new knowledge is acquired. However, the Five-Labs study’s vision of the role of technology development will not serve us well if the climate picture darkens, and it turns out we do need an effective long-term response to climate change.
1. The estimated range of impacts varies widely. See R. Prinn, H. Jacoby, A. Sokolov, C. Wang, X. Xiao, Z. Yang, R. Eckaus, P. Stone, D. Ellerman, J. Melillo, J. Fitzmaurice, D. Kicklighter, G. Holian, and Y. Liu, “Integrated Global System Model for Climate Policy Assessment: Feedbacks and Sensitivity Studies,” Climatic Change, forthcoming 1998.
2. For a lucid analysis of this distinction, using somewhat different terminology, see A.B. Jaffe and R.N. Stavins, “The Energy Efficiency Gap: What Does It Mean?” Energy Policy 22(10)(1994): 804-810.
3. Frequently referred to as the Five-Labs study, see Interlaboratory Working Group, “Scenarios of U.S. Carbon Reductions,” Interlaboratory Working Group on Energy-Efficient and Low-Carbon Technologies, Office of Efficiency and Renewable Energy, U.S. Department of Energy, 1997.
4. It is worth pointing out that the size of the needed reduction is subject to considerable uncertainty; see M.D. Webster, “Uncertainty in Future Carbon Emissions: A Preliminary Exploration,” Report No. 30, MIT Joint Program on the Science and Policy of Global Change, Cambridge, Mass., 1997. Our own MIT base implies a 470 MMtC reduction to achieve 1990 levels, and the latest Administration analysis uses a 2010 baseline which would yield a number in the same range; see U.S. Administration, The Kyoto Protocol and the President’s Policies to Address Climate Change: Administration Economic Analysis, July 1998. The larger the number, the higher the cost.
This paper was prepared for the September 23, 1998, policy conference sponsored by the ACCF Center for Policy Research, and will be published in the Center’s forthcoming book, Climate Change Policy: Practical Strategies to Promote Economic Growth and Environmental Quality.