304

Reflections—Energy Efficiency Policy: Pipe Dream or Pipeline to the Future? Tom Tietenberg∗

Introduction As the world grapples with creating the right energy portfolio for the future, energy efficiency policy is playing an increasingly prominent role. In recent years, the amount of both private and public money dedicated to promoting energy efficiency has increased a great deal. Some of this has resulted from performance mandates (including efficiency standards for appliances and automobiles), but other sources include subsidies financed by system benefit charges (typically imposed on consumers of electricity) or by the revenue derived from the sale of carbon allowances in regional auctions. Is this a good investment? In this “Reflections” article, I examine this question from both theoretical and empirical perspectives, focusing almost exclusively on issues of diffusion and adoption rather than innovation. I begin by examining the remaining potential for increased energy efficiency. This is followed by a discussion of the limitations of the market in promoting investments in cost-effective energy efficiency. Next I present the case for government intervention and discuss the types of policy instruments available. Finally, in the last two sections, I review the evidence on the effectiveness of the various forms of intervention and then summarize my findings and present some conclusions.

The Potential for Increased Energy Efficiency The role of energy efficiency in the broader mix of energy policies depends of course on how large the potential for energy efficiency actually is. Are there any remaining energy efficiency opportunities to exploit? High energy prices, by themselves, should have promoted private interest in energy efficiency, and, in fact, many public policies have already been implemented to stimulate energy efficiency (Gillingham et al. 2004). ∗ Upon his retirement from teaching in May 2008, the author immediately accepted an appointment as one of three Trustees of the Energy and Carbon Savings Trust. This Trust was established by the Maine legislature to receive all of the revenues from the sale of Maine carbon allowances under the Regional Greenhouse Gas Initiative (RGGI), a cap-and-trade policy covering ten states in the northeastern United States. The Trust is mandated to invest the proceeds in energy efficiency within the state, being guided by benefit–cost and cost-effectiveness criteria. RGGI may be the first program in the world to use this specific source of revenue to promote energy efficiency, but it is unlikely to be the last.

Review of Environmental Economics and Policy, volume 3, issue 2, summer 2009, pp. 304–320 doi:10.1093/reep/rep004 Advance Access publication on June 11, 2009  C The Author 2009. Published by Oxford University Press on behalf of the Association of Environmental and Resource

Economists. All rights reserved. For permissions, please email: [email protected]

Energy Efficiency Policy

305

Although estimating the remaining potential is not a precise science, the conclusion that significant opportunities remain seems common. For example, the eleven studies examined in Nadel et al. (2004) indicate that “a very substantial technical, economic and achievable energy efficiency potential remains available in the U.S.” Detailed studies that focus on particular sectors seem to come to a similar conclusion (Gellings et al. 2006). Using a survey approach, Sorell et al. (2004) found that “Interviewees were nearly unanimous in agreeing that there were many opportunities with paybacks of less than 3–4 years.” Given the nature of the externalities examined below, it is likely that the set of socially cost-effective opportunities is even higher than suggested by these estimates, which focus only on the private returns. The existence of so many as yet unexploited opportunities for investing in cost-effective energy efficiency has been referred to by some as the “energy efficiency paradox” (DeCanio 1998). Why haven’t these opportunities been exploited?

Barriers to Efficient Market Choices Clearly, the market provides incentives for energy efficiency, even in the absence of any government intervention. Higher energy prices naturally trigger efforts by both consumers and firms to reduce energy costs, including investing in energy efficiency. However, numerous barriers have discouraged or prevented investments in cost-effective energy efficiency opportunities, suggesting that government intervention may be necessary.

Externalities Not surprisingly, the presence of externalities provides one, but not the only, foundation for thinking about the role of government policy. In terms of types of externalities affecting the market level of energy efficiency, two seem particularly important in the current policy context: climate change and national security. • The combustion of fossil-fuel energy for heat and/or to produce electricity creates greenhouse gases, which, upon accumulating in the atmosphere, modify the climate. This climate modification is expected to impose considerable economic damage on the world economy, but most of these damages are external to any entity making an energy choice. The fact that these are external costs creates a bias away from energy efficiency, a bias that is larger for those fuels that contribute more greenhouse gases per unit of energy consumed (such as unsequestered carbon emissions from coal). • National security issues enter the mix because imported energy can be more risky, not only in terms of possible negative effects on the economy from an embargo, as was the case in the 1970s, but also in terms of the higher security risks posed by the use of oil revenues to finance any terrorist activities, either directly or indirectly (Friedman 2008, Chapter 4). For our purposes here, the point is that these risks are an externality to most consumers, who in any case typically have no idea where their fuel originates. This creates another bias against energy efficiency, since imported energy appears to be cheaper than it actually is (considering the risk premium), especially for the riskiest imports.

306

T. Tietenberg

In addition to these externalities, a host of organizational issues have been identified as contributing to the energy efficiency paradox described above. For the purposes of this review, the main barriers to capturing unexploited, cost-effective energy efficiency opportunities can be organized into five categories: information deficiencies, perverse incentives, limited availability of resources (especially capital), price volatility, and nonoptimizing behavior.

Information Deficiencies Information deficiencies that impede rational choice can result either from a general lack of knowledge (i.e., information that is either missing or hidden) or, more likely, because the information is asymmetrically available.

Missing or Hidden Information Missing or hidden information can involve either unrecognized associated costs or benefits. Costs or benefits can be unrecognized in the decision, for example, because they don’t surface until after the purchaser has had some experience with the product or service. New lighting or quieter motors, for example, can affect worker productivity. Energy efficient products might also turn out to have higher or lower operating or maintenance costs than expected. Associated costs and benefits that are identified only with experience are especially likely to remain unrecognized for those products that are either completely new or are experiencing rapid technological change. Rapid technological change depreciates previous experience simply because that experience may or may not be applicable to the new version. To the extent that the unrecognized benefits or costs are context specific (for example, they depend on the particular work environment or building design), the experience gained by early users may not necessarily be relevant for others. Context-specific information may not prove particularly helpful even if incentives are sufficient to make sure the information is widely shared. New or technologically different products face another problem as well. First-users, let’s call them pioneers, commonly face both higher reliability problems and higher unit costs than subsequent users. Higher reliability problems may be common because early products or services have not yet had the advantage of using market feedback to identify and correct problems. Higher costs can flow from an inability of the low initial production levels to take advantage of economies of scale. Both of these effects diminish the incentives for potential users to invest early, encouraging them instead to wait until the bugs are worked out and the costs fall.

Asymmetric Information Asymmetric information problems arise in situations where the information is available, but to someone other than the party making the energy investment decision. For example, the producer of a more energy efficient product or the supplier of an energy efficiency service may know that this product or service will pay for itself (in reduced energy costs) in a very short period of time, but potential purchasers may not.

Energy Efficiency Policy

307

In many cases, the market works quite effectively to reduce information problems. Sellers and suppliers certainly have an incentive to inform purchasers that their energy efficient products or services will ultimately save money. Information supplied by the market, however, may not always be sufficient or credible. Sellers have an incentive to overstate the benefits of their product if they can get away with it. Rapid technological change in this industry, coupled with the fact that many of the products are purchased infrequently, adds to the inability of purchasers to test the proffered information against their own direct experience. Role of Energy Service Companies One very interesting market response to this information problem is the rise of energy service companies. Some energy service companies contractually guarantee that cost savings will meet or exceed annual payments to cover all energy efficiency project costs—usually over a period of seven to 10 years. If the guaranteed savings don’t materialize, the energy service company pays the difference. In this case, the energy service company, rather than the customer, bears both the information burden and the risk. Given this allocation of risk, the information transfer becomes credible. It is important to note, however, that in the absence of government subsidies these companies usually only serve a portion of the market, since this strategy works best for large commercial or industrial enterprises. Bounded Rationality Another explanation for deficient information that has been advanced in the literature is the notion of bounded rationality (Dyner and Franco 2004). Originally introduced by H. A. Simon (1982), this concept suggests that since time, resources, or cognitive ability can limit the scope of human investigation of options, purchasers may not consider sufficient aspects of the problem to reach the optimum choice. Rather they “satisfice” or make the best choice under a limited set of considered options. The energy efficiency literature has found that bounded rationality does seem to characterize decision-making in some specific cases. For example, Stern (1984) noted that routines are rather commonly substituted for rigorous decision-making. These routines, such as replacing a depreciated piece of equipment with the same brand and type, may economize on the time and effort spent searching for the best product or strategy, but they can lead (and have led) to substantial biases against energy efficiency when technologies are rapidly changing. Interviews with managers have also demonstrated how common it is for the time available for investigating energy efficiency options to be severely limited (Sorell et al. 2004, p. 295).

Perverse Incentives Cost-effective energy efficiency investments can also remain unexploited because of perverse incentives. One classic example involves the relationship between landlord and tenant. The landlord commonly has the responsibility for making any energy efficiency investments to a rental unit (installing windows with a higher insulating factor, for example), but it is the tenant who commonly pays the energy bill. If the tenant had perfect information about the energy costs for each prospective unit he might rent, he would be willing to pay a higher rent

308

T. Tietenberg

for the energy efficient unit whenever the extra expense associated with the higher rent could be recovered through lower energy costs. In this case, the landlord has an incentive to make the investment and charge more rent to recover the cost. It is difficult, however, for any prospective tenant to compare annual energy costs across the units he is considering. Since, for example, energy consumption data depend on tenant behavior as well as the efficiency of the unit, prospective renters will tend to discount energy information supplied by the landlord (even if it is accurate, which is itself not obvious). Faced with this reality, the landlord’s incentive is to underinvest in energy efficiency because of the inability to charge a sufficiently high rent to recover the cost. Another classic example involves the situation where the consumer is not billed directly for energy use. One common reason for not directly billing the consumer is that the cost of establishing and monitoring individual heating zones (involving additional investments in both plumbing and monitoring) precludes separate monitored zones in these units. Familiar cases involve dorm rooms and rental units. Hotel rooms represent a somewhat different case, where the room may have a dedicated heat source or zone, but it is not practical to monitor individual use and to bill on such a short-term basis. In these cases where the individual faces no economic consequence from consumption choices, the tendency is for energy consumption to be inefficiently high, since the marginal cost of additional use is zero. In addition, it may not make sense for the principal to pay for the energy efficiency investment when the agent’s consumption behavior is likely to reduce the cost effectiveness of the investment. Consider, for example, a common occurrence in university dorms where, even during the winter, some windows are left open, either for ventilation or to lower the temperature in top-floor rooms as the heat rises and is trapped by ceiling insulation. Investing in energy efficient windows obviously would not have a high payoff if the windows were systematically left open.1 Even within companies, principal-agent problems may prevent the adoption of energy efficient strategies or products. One common example involves the organizational separation of responsibilities for capital and operating costs. In this situation, those managing new construction may be tasked with holding capital costs down, but get no credit for lowering operating costs. When the offsetting energy reductions are either not considered at all or receive less weight than they should in the investment decision, firms hurt their own interests by investing too little in energy efficiency.

Limited Availability of Capital Energy choices may be constrained by the limited availability of resources—capital in particular. In the typical case, the agent trades off the higher up-front cost of an energy efficient investment against the subsequent energy savings. When the investment has a good rate of return, the agent can simply borrow the money, using the savings to pay off the loan. Suppose, however, that for some reason the money cannot be borrowed or the availability of capital is sufficiently constrained so that it can be accessed only by paying a very high interest rate. Such examples are not hard to find. Municipalities face budget limitations, and 1

Note that a policy that subsidizes energy efficient windows would not solve this problem. As this example illustrates, for policy intervention to be effective, the context matters.

Energy Efficiency Policy

309

typically a high degree of spending, at least in the short run, is nondiscretionary. Low-income residents may find that their lack of collateral limits their borrowing choices. Even firms can face capital constraints that affect both internal and external sources of capital (Sorell et al. 2004, p. 298). As a recent World Bank publication (Taylor et al. 2008) points out, capital constraints make financing energy efficiency in developing countries particularly difficult: “As project opportunities tend to be relatively small scale and dispersed, transaction costs can prove daunting unless mechanisms are put in place to take advantage of similarities among projects and bundle them. Some form of financial intermediation is usually required, unless enterprises use their own funds. Typically, therefore, implementation of energy efficiency projects involves interaction of both financing entities and technical experts with clients. Project delivery requires very efficient contracting to achieve this without driving up transactions costs—a challenge in any country, but especially where market institutions may be relatively weak, causing greater insecurities in contracting, as in Brazil, China, and India.” (p. 3)

Price Volatility It has been well known for some time that investment is impeded by price volatility, particularly when those investments involve an element of irreversibility (see, e.g., Pindyck 1991). This, in combination with the rather large volatility of oil prices recently, suggests that price volatility may well be one source of the current energy efficiency paradox.

Nonoptimizing Behavior Economic models typically assume that decision-makers have full information and that they use this information to optimize outcomes. As we saw above, the full information assumption may be flawed, which has implications for both private and policy outcomes. The field of behavioral economics, which has brought research in psychology to bear on the science of decision-making, has also questioned the second assumption—namely that fully informed humans necessarily engage in optimizing behavior. This phenomenon has been referred to as “behavioral failure” in an earlier issue of this journal (Shogren and Taylor 2008). Viewing energy efficiency through the lens of behavioral failure can shed some additional insight not only on why even cost-effective energy efficiency opportunities remain unexploited, but also on why policy instruments such as information strategies frequently prove insufficient to promote even privately cost-effective outcomes. To illustrate this point, I focus on one main finding of the field, which has become known as the “status quo bias.” As noted by Samuelson and Zeckhauser (1988): “The main finding is that decision makers exhibit a significant status quo bias. Subjects in our experiments adhered to status quo choices more frequently than would be predicted by the canonical model.” (p. 8) As a practical matter, in our context this implies that consumers may not always choose the more energy efficient product or service even when, according to conventional logic,

310

T. Tietenberg

that choice appears to be in their best interest. This aspect of the energy efficiency paradox emanates from an asymmetry known as “loss aversion.” One implication of loss aversion is that individuals have a strong tendency to retain the status quo, because they perceive that the disadvantages of change outweigh the advantages (Kahneman et al. 1991). The policy implication is rather profound; contrary to conventional wisdom, internalizing all external costs and providing adequate information may not be sufficient to produce an efficient level of energy efficiency.

The Case for Government Intervention As Jaffe and Stavins (1994) point out in their classic article, the existence of unexploited, costeffective energy efficiency opportunities can be thought of as a necessary, but not sufficient, condition for government intervention. Sorell et al. (2004) provide a nice summary of these arguments and distinguish between sources of unexploited efficiency opportunities that require policy intervention and those that don’t. The nature of the barriers discussed above, which are responsible for the gap, affects the desirability of government intervention, the appropriate level of efficient government intervention, and its form. This section considers the desirability and level of intervention, leaving the more intensive discussion of its form to the next section.

Externalities The strongest case for government intervention flows from the existence of externalities. Markets are not likely to internalize these external costs on their own. As a result of the national security and climate change externalities, as well as other external co-benefits such as pollution-induced community health effects, energy efficiency is too low.

Information Deficiencies In terms of their role in justifying government intervention, information deficiencies are a mixed bag. If the perfect information, cost-effective outcome would be to invest in more energy efficiency, government intervention to overcome these information deficiencies could be justified, assuming the costs of intervention do not exceed the benefits. Not all information deficiencies merely camouflage good investment opportunities, however. Any hidden costs might, when added to the initial investment, exceed the benefits. Hidden costs do exist. For example, many people switching to compact fluorescent lights (CFLs) were chagrined to discover that, due to their mercury content, CFLs had special disposal requirements, frequently involving a fee, that did not apply to normal incandescent bulbs. In practice, however, the literature suggests that hidden benefits are much more significant than hidden costs. In their survey of firms, Sorell et al. (2004, pp. 292–93) found that the only hidden cost of any significance was the potential loss of production during equipment installation or upgrade. That specific cost would accompany any similar capital installation or upgrade; it is not unique to energy efficiency. In contrast, the literature seems to suggest that the number and types of hidden co-benefits are larger. Common benefits other than the targeted energy savings from energy efficiency

Energy Efficiency Policy

311

investments would include water savings, lower maintenance costs, and increased worker productivity. Notice that these all accrue to the decision maker. Once these benefits become apparent, they should provide a private incentive for an increase in energy efficiency investments. In this sense, they are quite different from the external benefits, mentioned above, which are not likely to be efficiently exploited in the absence of government intervention.

Time and Capital Constraints Both time and capital constraints could also justify government intervention. If time or manpower constraints serve to limit the availability of information, the government could supply information. Given the inefficiency of having every firm or household individually research all the options, thereby causing an enormous duplication of effort, having the government involved in the generation and dissemination of this information would seem to be a potentially efficient strategy. This is the rationale for the common government practice of identifying and disseminating “best practices” as well as labeling strategies, such as the U.S. Energy Star program for appliances, a subject investigated in more detail below. Even with an efficient flow of information, efficient investments can be prevented or limited by capital constraints. In this case, government intervention to increase the availability (and lower the cost) of capital can also promote an efficient increase in energy efficient investment. One underexplored policy avenue, at least in the United States, involves directly addressing the dampening effects of fossil fuel price volatility on energy efficiency investment. The closest parallel, apparently used very successfully in Germany to promote electrical generation from renewable resources, is the “feed in tariff.” Rather than subsidize the cost of the initial investment, this approach guarantees the prices received for electricity production over the lifetime of the investment (Michel 2007). Whether and how this approach might apply to energy efficiency investments remains to be seen, but it is an intriguing idea.

Nonoptimizing Behavior The final category, nonoptimizing behavior, is particularly troubling since it suggests that the normal policy prescriptions of internalizing the externalities, assuring an adequate supply of reliable information, and lowering the capital constraints may not be enough. In this case, deeper intervention might well be needed to overcome the status quo bias as long as the investment expenses plus the administrative and transactions costs associated with the intervention did not exceed the gains from the energy efficient investment. Although the effectiveness of government intervention in overcoming nonoptimizing behavior remains to be seen, a deeper exploration of this intriguing question remains to be answered by future research. One useful observation about the barriers to cost-effective energy efficiency, noted in an extensive survey by Sorell et al. (2004), is the frequent occurrence of multiple limiting barriers for a single energy efficient investment. The practical implication of this important insight is that limiting government intervention to one specific instrument may be insufficient, since it could simply cause another barrier to become the binding constraint. Thus, government intervention may need to invoke multiple instruments.

312

T. Tietenberg

Forms of Government Intervention Governments that have decided to intervene in energy efficiency markets have a number of potential policy instruments at their disposal. Different instruments may have very different consequences, not only in terms of their effectiveness in enhancing energy efficiency, but also in terms of their impact on such related areas as climate change and import dependence (Jaffe and Stavins 1995).

Internalizing Externalities The most efficient way to address the climate change and national security externalities is with instruments targeted specifically to those problems. If those targeted policies have not been implemented, some role for policies focusing strictly on energy efficiency could be justified in order to counteract the derived bias against energy efficiency. In the case of climate change, the targeted policy would normally be either a tax on all greenhouse gases or a cap-and-trade system. By raising the costs associated with consuming the energy responsible for the emissions, this policy would eliminate the relative bias against fuels with little or no contribution to climate change and would promote more energy efficiency. There are important differences between the tax and cap-and-trade approaches. However, these differences have been discussed recently in this journal (Goulder and Parry 2008; Metcalf 2009; Keohane 2009; and Murray et al. 2009), and are therefore not repeated here. The most efficient way to address the national security bias toward imports would involve a tariff on imports that incorporates the risk premium. This policy would internalize the externality associated with vulnerable imports, raising the costs of those imports and providing an efficient incentive for energy efficiency projects. This could lower national security costs by lowering the dependence on those imports.

Information Strategies Information barriers can be addressed either by providing more information or by eliminating choices that are clearly inefficient. A number of different instruments can be used. Credible information about products can be provided through certified labels, with the certification provided by the government (as in the Energy Star program), or by independent private organizations, such as the Leadership in Energy and Environmental Design (LEED) certification for buildings. Energy Star, a joint program of the U.S. Environmental Protection Agency and the U.S. Department of Energy (DOE), provides information to both consumers and businesses. Products in more than fifty categories, ranging from appliances such as refrigerators, dishwashers, and room air conditioners to heating and cooling options such as heat pumps, furnaces, and programmable thermostats, can qualify to apply the Energy Star label if they meet the certification performance standards. LEED is a third-party certification program that has established standards for a number of aspects of the design, construction, and operation of high-performance green buildings in the United States. Points are assigned for meeting or beating predetermined levels of performance in five key areas: sustainable site development, water savings, energy efficiency, materials

Energy Efficiency Policy

313

selection, and indoor environmental quality. Four levels of green building certification— Certified, Silver, Gold, and Platinum—are awarded based on the total points earned.

Minimum Efficiency Standards Although they provide additional information, labeling programs leave the level of energy efficiency to the purchaser. In a context of bounded rationality or nonoptimizing behavior, however, other strategies such as performance standards or mandates may prove more effective in achieving cost-effective outcomes (Sanstad and Howarth 1994). Minimum or average energy efficiency standards for buildings and products eliminate inefficient choices from the feasible set of alternatives. In the United States, the Appliances and Commercial Equipment Standards Program promulgates minimum efficiency standards for residential appliances and commercial equipment. These minimum standards apply to all major home appliances sold in the United States, regardless of whether they were domestically manufactured or imported. Manufacturers must also use standard test procedures developed by DOE to prove the energy use and efficiency of their products. Test results are communicated to consumers via a yellow Energy Guide label, which manufacturers are required to display. This label estimates energy use and compares it to that of similar products, and even lists approximate annual operating costs. While manufacturers must discontinue manufacturing products that do not meet the efficiency standards, products manufactured before the effective date of the new standards may still be sold. Because U.S. federal appliance standards limit choice, Sutherland (1991) described them as causing a welfare loss that places a particularly heavy burden on poor families. A subsequent analysis by Soft (1993) found that this loss, to the extent it exists at all, is very small. It is important to remember, however, that it is not only the type of policy instrument that matters, but also its level. Jaffe and Stavins (1995), for example, suggest that their finding that building codes did not seem to have much impact on the diffusion of thermal insulation might well have been due to the fact that the minimum energy efficiency standards in the codes were typically lower than the prevailing norm and, hence, nonbinding.

Transferable White Certificates The newest approach to government intervention in energy efficiency markets in Europe involves the use of transferable white certificates (TWC) (Langniss and Praetorius 2006). This approach, which mirrors other tradable permit schemes, requires a specified amount of energy efficiency to be achieved from stipulated parties. The parties demonstrate compliance by turning in the required number of white certificates, denominated, for example, in kilowatthours saved, at the end of the compliance period. Certificates can be earned by engaging directly in demonstrated energy saving activities or by acquiring certificates from others who have saved more energy than necessary to achieve compliance. In the past few years, France, Italy, and Great Britain have embarked on implementing TWC schemes, while other European Union Member States (e.g., the Netherlands, Denmark, and Poland) are in the initial stages of investigating and/or implementing TWC schemes (Mundaca 2008). Since these programs are still in their early stages, the literature provides little information about how successful they are in practice.

314

T. Tietenberg

One characteristic that does not seem generally recognized, but may turn out to be important, is the effect of TWC schemes on “cream skimming.” Many current subsidy programs targeted at energy efficiency attempt to obtain the most reduction per dollar expended—picking the lowest hanging fruit. However, the competitive environment that seeks to demonstrate cost-effective results may overlook collateral energy efficient opportunities, even when these opportunities could be exploited much more cheaply if they were funded at the same time as the primary opportunities. For example, due to the fixed cost of getting all the manpower and equipment to the site, making all cost-effective investments at the time of the first visit can be much cheaper than requiring multiple sequential visits to handle the remaining opportunities. Nonetheless, in a competitive environment for funding, including all opportunities in a single funding proposal can make that proposal seem less cost effective, thereby putting it at a disadvantage with a proposal that considers only the most cost-effective opportunity. A TWC program directly confronts this deficiency by mandating the responsibility to achieve the efficiency target, not just picking the low hanging fruit. As such it provides incentives for firms not only to invest in a broader range of opportunities, but also to time those investments to minimize the present value of the costs.

Subsidies As noted in the opening paragraphs of this article, the increased flow of public funds into the market for energy efficiency has led to an increase in the use of subsidies. Subsidies attempt to reduce the number of unexploited energy efficiency opportunities by making energy efficiency investments cheaper for the private investor, simply by having the public sector absorb some proportion of the cost. Lowering the supply curve to private investors should, all other things being equal, mean more investment. The most common historic subsidy for electric energy efficiency was funded by the use of a small mandatory per-kilowatt-hour charge (typically called a “system benefit charge” or “public benefit charge”) attached to the distribution portion of the customer’s bill (Nadel and Kushler 2000). The services funded by these programs range from educational initiatives and funding for low-income customers to environmental and efficiency programs. Prior to the restructuring of the electricity sector in the United States, the administration, design, and delivery of ratepayer-funded energy efficiency program activities had largely been the responsibility of the regulated utilities. Many states that restructured their electricity sectors have now established alternative structures under which program administration and governance have been taken over by nonutility entities, such as existing state governmental agencies or nonprofit corporations with boards of directors and state oversight (Blumstein et al. 2005). The newest source of funding for energy efficiency subsidies is the revenue accrued from carbon cap-and-trade programs. Whereas cap-and-trade programs have historically provided allowances to emitters for free, a number of states have recently established programs where the allowances are being (or will be) auctioned off with proceeds used to fund energy efficiency. The motivation for the use of subsidies funded in this manner is in no small part an effort to reduce the financial burden on consumers as the costs of carbon control are passed on by the emitters; by reducing consumption, energy efficiency helps cushion the shock from the higher prices.

Energy Efficiency Policy

315

As useful as they may be in this regard, subsidies pose some challenges of their own. First, depending on how the subsidy program is designed, it is possible that public money could simply, in the limit, displace private money, resulting in a shift in the distributional burden of the cost, but no change in the total level of investment in energy efficiency. Recognizing this potential, most subsidy programs require a degree of cost sharing. One common form is to have the public sector pick up whatever proportion of the investment is necessary to lower the payback period to a specific target (say, 2–4 years). Another form, the one used in Maine under the Regional Greenhouse Gas Initiative (RGGI) program, is to allocate the money on the basis of a competitive bid, where electricity energy efficiency bids are ranked on the basis of kilowatt-hours saved per public dollar allocated to the program. Using public dollars (rather than the sum of public and private expenditures) creates an incentive for applicants to cost share. Under this set of rules, all other things being equal, cost sharing is more likely because bidders that contribute a larger share of the total cost are more likely to be funded. Another important challenge concerning the use of subsidies is posed by what has become known as the “rebound effect.” According to the rebound effect, an energy efficiency measure can trigger behavioral reactions that lead to an increase in the service now more efficiently supplied. Thus, while the investment would cause energy input (and associated emissions) per kilowatt-hour to decline, the resulting lower cost of kilowatt-hours would provide an incentive for consumers to consume more of them. Depending on its magnitude, this offsetting increase in kilowatt-hours would reduce and could even negate the energy input and emissions savings brought about by the energy efficiency investment. As Berkhout et al. (2000) point out, a change in energy efficiency can trigger three different rebound effects: a direct price effect, a real income effect at the micro level, and a composition of demand effect at the macro level. Whereas the first two effects unambiguously lead to more electricity consumption, the sign of the third effect is ambiguous. Several studies provide some guidance on the magnitude of these effects: • Focusing on empirical results from the Netherlands, Berkhout et al. (2000) conclude that “according to every definition, empirical evidence shows that the [rebound effect] is probably small: between 0 and 15 percent.” • A contemporaneous survey of studies focusing primarily on the United States (Greening et al. 2000) found that “in the majority of cases, technical efficiency gains result in fuel savings, which are only slightly eroded by increases in demand.” (p. 399) • In one of the few general equilibrium studies able to capture the macro effects, Grepperud and Rasmussena (2004) examined the effects in Norway and found different magnitudes of rebound effects within manufacturing. Specifically, they found that rebound effects appear to be most important for the manufacture of metals, somewhat less important for the chemical and mineral products sector, and least important for pulp and paper.

Evaluation of Government Policies and Programs Policies can be evaluated using ex ante or ex post studies. Ex ante studies forecast what effects the program could be expected to have if implemented, and are frequently used to determine

316

T. Tietenberg

the desirability of the program or to understand the consequences of certain programmatic design features. Ex post studies attempt to decipher how well the program worked in practice. Not surprisingly, most of the studies are of the ex ante variety. However, this section focuses mainly on those studies that evaluate what actually happened.

Early Evaluations In the early 1990s, ex post evaluations of energy efficiency policy produced more controversy than consensus, particularly in terms of how much of the energy efficiency paradox could be resolved and at what cost. Joskow and Marron (1993) combined reported outcomes with some program projections from a sample of ten major utilities to compare the actual cost per kilowatt-hour saved in utility conservation programs to the supply curves projected by earlier analyses. They found costs to be significantly higher than would be suggested by the supply curves. This led to a spirited defense of the earlier analyses by Amory Lovins (1994), a principal author of these analyses. Two important considerations about this dispute should be kept in mind when assessing the value of current energy efficiency policy. First, as Joskow and Marron note, the dispute was one of degree, not one that cast doubt on the overall desirability of energy efficiency policy. Second, a lot of water has gone over the dam since these analyses were conducted. Electricity prices have increased, information has improved, more skilled labor is now available in the energy efficiency field, utility incentives have been changed by decoupling programs,2 and new governance structures for delivering energy efficiency have been established, in part as a response to past inefficiencies (Blumstein et al. 2005).

Macro Evaluations Energy efficiency policy can be evaluated at the macro economy level or the programmatic level. Here we examine the more macro studies, while we discuss those at the more programmatic level below. Unfortunately, in this area as in many others in environmental and natural resource economics, the number of credible ex post analyses is rather thin (Tietenberg and Johnstone 2004). In a rather unique, but very interesting, approach, Joskow (2003) conducted an ex post evaluation of two highly influential energy forecasts that focused on the year 2000 and were published in 1979 (Landsberg et. al. 1979; Schurr et al. 1979). He discovered, among other things, that both forecasts overestimated energy consumption by about 15–20 percent. Joskow suggests that the most likely explanation for what occurred in the residential sector was a combination of policy driven energy efficiency improvements and changes in the economic mix (from more energy using to less energy using products), while in the industrial and commercial sectors policy driven energy efficiency seemed to dominate. This finding has some support from other studies involving very different methodologies. For example, 2

Because they reduce kilowatt-hour sales, efficiency programs can be harmful to a distribution company’s financial performance. One solution is to “decouple” utility sales from revenues so as to reduce or eliminate any disincentive for utilities to invest in energy efficiency.

Energy Efficiency Policy

317

Geller et al. (2006) found that energy efficiency policies reduced energy consumption by 11 percent. An econometric analysis of a cross-section of states by Berry (2008) confirms the expectation that states with higher utility efficiency program expenditures per capita and a greater range of other efficiency programs experienced the greatest reduction in the growth of power sales. Furthermore, annual reports for the state organizations implementing energy efficiency uniformly report that these savings are quite cost effective, with per unit costs well below the cost of generating the electricity. Further confirmation is provided by a close examination of California’s aggressive program, which has been described as “the most successful statewide energy conservation campaign in history” (Bachrach 2003, p. 37). California’s electricity use per person increased only slightly over the past quarter century, compared to a 50 percent per person increase for the rest of the nation (Bachrach 2003, Figure 2).

Programmatic Evaluations In addition to evaluations performed at the national and state levels, there have also been ex post evaluations of specific programs. We begin with energy research and development (R&D). A comprehensive National Research Council review of federal R&D efforts (COB 2001) examined seventeen R&D programs in energy efficiency funded by DOE. The study estimated that the total realized economic benefits associated with the energy efficiency programs (valued in 1999 dollars) were in the $30 billion range, while the costs were roughly $7 billion over the 22-year life of the programs. Sanchez et al. (2008) quantified the energy and carbon savings from the U.S. Energy Star program and identified the specific portions of the program responsible for the lion’s share of the savings—office equipment products, including monitors, computers, and imaging equipment. No estimates were provided of program costs. An earlier article by Howarth et al. (2000) identified a key source of the success for the office equipment portion of the Energy Star program. Not only did the label provide important information to energy-conscious consumers, but an executive order specifying that government agencies buy Energy Star office equipment whenever possible greatly expanded the market for these products. Motivated by the desire to participate in this expanded market, firms opted to meet the Energy Star requirements for the great majority of their products. This combination of a voluntary program that provided better information to consumers, coupled with a program of strong incentives for firms to upgrade their products in order to participate, apparently did a better job of overcoming multiple barriers than either program would have done alone. Of special relevance to the current increase in funds available for adoption subsidies, the literature suggests that such subsidies may work better in practice than our theory would suggest. One telling example is the careful empirical examination of the diffusion of thermal insulation in new home construction by Jaffe and Stavins (1995). They found that, contrary to conventional wisdom, the likely effects of adoption subsidies on technology diffusion appeared to be substantially greater than the expected impacts of equivalent Pigouvian taxes.

318

T. Tietenberg

More prescriptive policies also seem to fare well when examined closely. Gillingham et al. (2004), for example, found appliance standards to be cost effective, with positive net benefits from energy savings alone and additional benefits from ancillary reductions in air pollution. This is particularly interesting because it reinforces the notion that neither higher energy prices nor providing information is sufficient in the face of bounded rationality and nonoptimizing behavior.

A Summing Up This “Reflections” review has found that considerable opportunities still exist to achieve costeffective savings via energy efficiency policy. From this perspective, energy efficiency policy clearly represents more of a pipeline to the future than a pipe dream. While this information provides a compelling basis for energy efficiency policy, we must recognize that the most detailed information that supports this view is focused mostly on electricity (and to a lesser extent natural gas), simply because that is where we have had the most experience. With the one exception of the corporate average fuel efficiency standards, the literature tells us much less about the effectiveness of policy in promoting energy efficiency in the direct combustion of oil for heating and transportation. That task remains for future research. The evidence presented here also seems to suggest that the challenge of breaking down the barriers to increased investments in energy efficiency is more difficult than previously thought. The message is not that the significant challenges posed by these barriers doom energy efficiency policies to failure, but rather that policy actions must recognize an expanded set of barriers and respond with some ingenuity in applying the available instruments. Information strategies, while helpful, are not enough. Typical energy efficiency policy strategies are too poorly targeted to respond effectively by themselves to both climate change and national security concerns. Complementary national policies such as carbon cap-and-trade, carbon taxes, and/or risk-targeted tariffs on oil imports would be necessary. But in the absence of these complementary policies, energy efficiency policies are still better than nothing. The perfect should not become the enemy of the good. Furthermore, the literature makes it quite clear that any second-best policy mix in the face of bounded rationality, nonoptimizing behavior, and volatile prices must recognize a role for more prescriptive strategies, such as targeted subsidies, efficiency standards, and/or tradable white certificates.

References Bachrach, Devra. 2003. Energy efficiency leadership in California: Preventing the next crisis. Electricity Journal 16(6): 37–47.

Berry, David. 2008. The impact of energy efficiency programs on the growth of electricity sales. Energy Policy 36(9): 3620–25.

Berkhout, Peter H. G., Jos C. Muskens, and Jan W. Velthuijsen. 2000. Defining the rebound effect. Energy Policy 28(6–7): 425–32.

Blumstein, Carl, Charles Goldman, and Galen Barbose. 2005. Who should administer energyefficiency programs? Energy Policy 33(8): 1053–67.

Energy Efficiency Policy Committee on Benefits of DOE R&D on Energy Efficiency and Fossil Energy, National Research Council (COB). 2001. Energy research at DOE—Was it worth it? Energy efficiency and fossil energy research 1978 to 2000. Washington, DC: National Academy Press. DeCanio, S. J. 1998. The efficiency paradox: Bureaucratic and organizational barriers to profitable energy-saving investments. Energy Policy 26(5): 441–54. Dyner, I., and C. J. Franco. 2004. Consumers’ bounded rationality: The case of competitive energy markets. Systems Research and Behavioral Science 21(4): 373–89. Friedman, Thomas L. 2008. Hot, flat and crowded: Why we need a green revolution and how it can renew America. New York: Farrar, Straus and Giroux. Geller, Howard, Philip Harrington, Arthur H. Rosenfeld, Satoshi Tanishima, and Fridtjof Unander. 2006. Policies for increasing energy efficiency: Thirty years of experience in OECD countries. Energy Policy 34(5): 556–73. Gellings, C. W., Greg Wikler, and Debyani Ghosh. 2006. Assessment of U.S. electric end-use energy efficiency potential. Electricity Journal 19(9): 55–69. Gillingham, Kenneth, Richard G. Newell, and Karen Palmer. 2004. Retrospective examination of demand-side energy efficiency policies. Resources for the Future, Inc. Discussion Paper 04–19. June, rev. Sept.). Goulder, Laurence H., and Ian W. H. Parry. 2008. Instrument choice in environmental policy. Review of Environmental Economics and Policy 2(2): 152–74. Greening, Lorna A., David L. Greene, and Carmen Difiglio. 2000. Energy efficiency and consumption—the rebound effect—a survey. Energy Policy 28(6–7): 389–401. Grepperud, Sverre, and Ingeborg Rasmussena. 2004. A general equilibrium assessment of rebound effects. Energy Economics 26(2): 261–82.

319 Jaffe, A. B., and R. N. Stavins. 1995. Dynamic incentives of environmental regulations: The effects of alternative policy instruments on technology diffusion. Journal of Environmental Economics and Management 23(3): S43–63. Joskow, Paul B., and Donald B. Marron. 1993. What does utility-subsidized energy efficiency really cost? Science 260(5106): 281–370. Joskow, Paul L. 2003. Energy policies and their consequences after 25 years. Energy Journal 24(4): 17–49. Kahneman, D., J. L. Knetsch, and R. H. Thaler. 1991. Anomalies: The endowment effect, loss aversion, and status quo bias. Journal of Economic Perspectives 5(1): 193–206. Keohane, Nathaniel O. 2009. Cap and trade, rehabilitated: Using tradable permits to control U.S. greenhouse gases. Review of Environmental Economics and Policy 3(1): 42–62. Landsberg, Hans, et al. 1979. Energy: The next twenty years, a report sponsored by the Ford Foundation and Administered by Resources for the Future (Cambridge: Ballinger). Langniss, Ole, and Barbara Praetorius. 2006. How much market do market-based instruments create? An analysis for the case of ‘‘white’’ certificates. Energy Policy 34(2): 200–11. Lovins, Amory. 1994. Apples, oranges, and horned toads: Is the Joskow & Marron critique of electric efficiency costs valid? Electricity Journal 7(4): 29–49. Metcalf, Gilbert E. 2009. Designing a carbon tax to reduce U.S. greenhouse gas emissions. Review of Environmental Economics and Policy 3(1): 63–83. Michel, Jeffrey H. 2007. The case for renewable feed-in tariffs. Journal of EUEC 1, http://www.euec. com/journal/articles/Article_V_01_I_01.htm. Mundaca, L. 2008. Markets for energy efficiency: Exploring the implications of an EU-wide “tradable white certificate” scheme. Energy Economics 30(6): 3016–43.

Howarth, Richard B., Brent M. Haddad, and Bruce Paton. 2000. The economics of energy efficiency: Insights from voluntary participation programs. Energy Policy 28(6–7): 477–86.

Murray, Brian C., Richard G. Newell, and William A. Pizer. 2009. Balancing cost and emissions certainty: An allowance reserve for cap-and-trade. Review of Environmental Economics and Policy 3(1): 84–103.

Jaffe, A. B., and R. N. Stavins. 1994. The energy-efficiency gap—What does it mean. Energy Policy 22(10): 804–10.

Nadel, Steven, and Marty Kushler. 2000. Public benefit funds: A key strategy for advancing energy efficiency. Electricity Journal 13(8): 74–84.

320 Nadel, Steven, Anna Shipley, and R. Neal Elliott. 2004. The technical, economic and achievable potential for energy-efficiency in the U.S.: A meta-analysis of recent studies. Proceedings of the 2004 ACEEE Summer Study on Energy Efficiency in Buildings, American Council for an Energy-Efficient Economy. Pindyck, Robert S. 1991. Irreversibility, uncertainty, and investment. Journal of Economic Literature 29(3): 1110–48. Samuelson, William, and Richard Zeckhauser. 1988. Status quo bias in decision making. Journal of Risk & Uncertainty 1(1): 7–59. Sanchez, Marla C., Richard E. Brown, Carrie Webber, and Gregory K. Homan. 2008. Savings estimates for the United States Environmental Protection Agency’s ENERGY STAR voluntary product labeling program. Energy Policy 36(6): 2098–108. Sanstad, A. H., and R. B. Howarth. 1994. “Normal” markets, market imperfections and energy efficiency. Energy Policy 22(10): 811–18. Schurr, Sam H., J. Darmstadter, H. Perry, W. Ramsay, and M. Russell. 1979. Energy in America’s future: The choices before us, a study by the staff of the RFF national energies project. Baltimore, MD: Johns Hopkins University Press. Shogren, Jason F., and Laura O. Taylor. 2008. On behavioral-environmental economics. Review of

T. Tietenberg Environmental Economics and Policy 2(1): 26–44. Simon, Herbert A. 1982. Models of Bounded Rationality, vols. 1 and 2. Cambridge, MA: MIT Press. Soft, Steven. 1993. Appliance standards and the welfare of poor families. Energy Journal 14(4): 123–28. Sorell, Steve, Eoin O’Malley, Joachim Schleich, and Sue Scott. 2004. The economics of energy efficiency: Barriers to cost-effective investment. Cheltenham, UK: Edward Elgar. Stern, P. C. 1984. Energy use: The human dimension. New York: W. H. Freeman. Sutherland, Ronald J. 1991. Market barriers to energy-efficiency investments. Energy Journal 12(3): 15–34. Taylor, Robert P., Chandrasekar Govindarajalu, Jeremy Levin, Anke S. Meyer, and William A. Ward. 2008. Financing energy efficiency: Lessons from Brazil, China, India, and beyond. Washington, DC: The World Bank. Tietenberg, T., and N. Johnstone. 2004. The management of ex post public policy evaluations. In Tradable permits: Policy evaluation, design and reform, ed. T. Tietenberg and N. Johnstone, 173–89. Paris: Organization for Economic Co-operation and Development.