While the rest of this report examines the prospects for increasing the supply of electricity from increasingly clean sources, this chapter deals with economically efficient methods of reducing electricity usage through energy efficiency.1 Energy-efficiency measures offer the promise of reducing energy use and saving money on electricity bills, as well as reducing negative environmental externalities associated with the production of electricity. These measures can provide pricing and usage transparency to allow residential, commercial, and industrial customers greater control of their energy choices. Appropriate federal and state policies can promote the development of more energy-efficient buildings and products and strengthen incentives for consumers, businesses, and industrial customers to pursue cost-effective energy-efficiency measures and to make investments that will provide future energy-efficiency improvements.
The improvements in energy efficiency achieved over the past 30 years can be attributed to a variety of factors, including technological progress and pressures on households and businesses to cut their spending on energy. In addition to those market forces, there is evidence that policies and programs designed to improve energy efficiency, such as energy-efficiency standards, funding for research and development (R&D), educational and informational efforts, and financial incentives to accelerate the development and adoption of energy-efficiency measures, have contributed to the improvement in energy efficiency experienced in OECD countries (Geller and Attali, 2005; citing
1 This chapter focuses on improving the efficiency of energy consumption. An improvement in energy efficiency occurs when there is a reduction in the energy inputs required to provide a given unit of energy services (e.g., lighting, cooling, heat, or drive power) to a specific end-user of such services. However, improvements in energy efficiency can occur with simultaneous increases in the use of energy services and in overall energy use.
Despite the large potential to reduce spending on electricity, however, market imperfections can lead to underinvestment in energy efficiency. One such market failure is attributable to unpriced pollution from the production of electricity, discussed in Chapter 2. But even if electricity is priced to include environmental externalities, there may be other market failures that lead to underinvestment in energy efficiency. For instance, consumers often have inaccurate or incomplete information about their energy use, price, or the net savings from energy-efficient investments. Asymmetric information or misaligned incentives, such as between landlord and tenant, may also reduce energy-efficient purchases. Improving information available to consumers through energy labeling, together with advanced metering infrastructure and customer systems, could increase cost-effective energy efficiency (NAS et al., 2010, NRC, 2011).
In addition to market failures, electricity consumers do not always behave in an economically rational way to minimize costs or maximize profits or benefits, and policies designed to reduce market failures may not affect behavior. Consumers may be reluctant to make new purchases of energy-efficient appliances because of inertia, risk aversion, or uncertainty about other characteristics of the appliances (see Wilson and Dowlatabadi  for a broad range of influences on consumer behavior). For these reasons, firms cannot completely capture the benefits of their investments in energy-efficiency innovations. This fact, coupled with the knowledge spillovers that impede realizing the full benefits that flow from their R&D (see Chapter 3), results in firms underinvesting in innovation in these technologies.
One challenge for policy makers, then, is to ensure that the benefits of energy-efficiency measures are greater than their costs. The savings from energy-efficient investments are difficult to calculate, however, and estimates derived from currently available methods have large uncertainty. Consequently, this chapter presents no estimates of the costs per kilowatt hour (kWh) of energy-efficiency measures. Improving the accuracy of savings estimates and measuring actual savings could aid in evaluating the effectiveness of existing and the design of future policies.2 Also beneficial would be understanding barriers that contribute to the gap between actual savings and costs and consumers’ product use, and developing behavioral models of how gains from energy efficiency can be realized (Allcott and Greenstone, 2012).
This chapter details potential electricity savings through energy efficiency, barriers to achieving the development and adoption of cost-effective energy-
2 It should be noted that there are other, unobservable costs and benefits to energy-efficiency investments that make their net benefits difficult to measure even if energy savings are correctly assessed (Allcott and Greenstone, 2012, p. 5).
efficiency technologies, and policies of the federal and state governments designed to address those barriers.
Americans today spend almost $400 billion annually on electricity to power their homes, offices, and factories, a large share of which is used in residential and commercial buildings (more than two-thirds of all the nation’s electricity use and 40 percent of total energy consumption) (EIA, 2015g [data as of April 2016]; see also Austin, 2012). On average across the United States, energy usage in buildings contributes about 41 percent of total U.S. carbon dioxide emissions (DOE, n.d.-a). And in some medium to large U.S. municipalities, building energy use can be responsible for 50 to 75 percent of citywide carbon emissions—much higher than the national average (City Energy Project, 2014, slide 9).3
As described in detail below, evidence suggests that energy-efficiency measures have been effective at reducing energy consumption. Moreover, there is potential for improvements in energy efficiency in the future. According to a previous Academies study, “Energy-efficient technologies for residences and commercial buildings, transportation, and industry exist today, or are expected to be developed in the normal course of business that could potentially save 30 percent of the energy used in the U.S. economy, while also saving money” (NAS et al., 2010, p. 278).4
Over the past 40 years, ongoing technological change, structural changes in the economy, changes in energy prices, and improved efficiency of energy use have contributed to the decline seen in energy intensity (units of energy as a share of per capita gross domestic product [GDP]) (NAS et al., 2010). Indeed, research suggests that three-quarters of the decline in U.S. energy intensity from 1970 to 2001 is attributable to improvements in the use of energy (Huntington, 2009; Levinson, 20155; Metcalf, 2008). All told, total energy used per dollar of goods produced is down, as is spending on energy services (from lighting to refrigeration) (EIA, 2016h). In sum, efficiency measures in the electricity sector
3 Using data from individual cities, it was found that buildings in the following cities contributed 50 to 75 percent of citywide carbon dioxide emissions, based on the cities’ individual greenhouse gas inventories: Dallas, Philadelphia, Minneapolis, Chicago, Washington (DC), New York City, and Salt Lake City.
4 Note that there is considerable uncertainty around the magnitude of energy-efficiency savings, particularly the costs of those improvements, discussed in detail in the rest of this chapter.
5 Levinson shows that 90 percent of pollution reduction in the manufacturing sector is due to changes in production technique rather in than manufacturing composition.
have reduced energy consumption and could continue to save energy in the coming decades, while also helping to reduce pollution (EIA, 2016h).
According to the Energy Information Administration (EIA), the adoption of energy-efficiency measures such as more efficient equipment, better insulation, and improved windows has contributed to a decline in residential energy use, as have migration patterns within the United States as more people have moved to states that are warmer in the winter (EIA, 2013b). Despite being 30 percent larger in size, newer houses use only 2 percent more electricity than those built before 2000 (EIA, 2009b, Forms EIA-457 A and C-G). Improvements in data on energy use due to technological advances in information collection, storage, and access also have contributed to improvements in energy efficiency in all sectors.
There is a literature investigating whether and why firms and consumers leave profitable or cost-effective energy-efficiency investments on the table. This “energy-efficiency gap” literature compares actual energy savings with the costs of energy-efficiency investments to see whether there are really unexploited energy-efficiency opportunities (Allcott and Greenstone, 2012; Jaffe and Stavins, 1994; Jaffe et al., 2004, Gillingham et al., 2009). If there is no energy-efficiency gap or the gap is negative, it means that the costs of the energy-efficiency measure outweigh the benefits.
Some researchers have questioned the existence or size of the energy-efficiency gap, offering evidence that predicted savings from certain energy-efficiency programs are overstated (Allcott and Greenstone, 2012; Davis et al., 2014; Dubin et al., 1986; Fowlie et al., 2015; Geller and Atalli, 2005; Gillingham and Palmer, 2014; Gillingham et al., 2009; Houde and Aldy, 2014; Jacobsen and Kotchen, 2013; Levinson, 2014a; Metcalf and Hassett, 1999).6 These studies highlight the need for rigorous, randomized evaluations of programs to ensure that those who do and do not receive a program or treatment are statistically identical. Allcott and Greenstone (2012) note that results on a small group of volunteers rather than a randomly selected group may be biased. Ramos and colleagues (2015, p. S19) note that “experimental methodologies with rigorous design and the use of large-scale random samples are extensively being employed to study novel aspects of energy efficiency in buildings.”
6 Other studies have found that models have overestimated the costs of energy-efficiency standards because of technological changes (Dale et al., 2009; Taylor et al., 2015). In addition, in a recent study, Kotchen (2016) finds that engineering forecasts do not significantly overestimate realized savings of energy efficiency, and Auffhammer and colleagues (2008) find that average savings and costs may be consistent with average utility-reported savings on demand-side management programs.
Much of the difference between energy savings predicted by current engineering models and actual energy savings may be attributable to how users adopt and use technology. Differences in consumers’ behavior may arise from differences in their values (including how much they discount future payments relative to the present) and their abilities to process or calculate savings from the available information. Another cause may be consumers’ increased use of an appliance when the costs of its use decline because of energy efficiency—termed the “rebound effect” (see, e.g., Gillingham et al., 2009, 2015). Recent empirical evidence suggests that the rebound effect in energy efficiency is small (Davis, 2008; Dumagan and Mount, 1993; Fowlie et al., 2015; but see Davis et al., 2014), but more analysis of this phenomenon is needed to understand how large it may be (see, e.g., Chitnis and Sorrell, 2015 [suggesting that including indirect income effects makes the rebound effect larger relative to relying only on price effects]). Finally, energy-efficiency investments may have hidden costs or large transaction costs that are not accounted for in engineering models (Fowlie et al., 2015).
To evaluate the cost-effectiveness of policies aimed at overcoming barriers to investment in energy-efficiency measures, it will be necessary to have better information on actual energy savings and market adoption rates of different technologies. Opportunities to collect such information have been enhanced by advances in communication and information technology. Analysts have hypothesized that a number of market imperfections, discussed below, inhibit customers from adopting energy-efficiency measures even when their benefits outweigh their costs, and the United States and Europe have adopted a number of measures designed to overcome these imperfections.
Economic, institutional, and political barriers to the development and adoption of energy-efficiency measures include
- incorrect energy prices (prices that do not reflect the full societal cost of energy, including costs of pollution), which reduces incentives to invest in energy efficiency (see Chapter 2);
- inadequate and imperfect information and—as anyone who has rushed to replace a broken water heater, furnace, or refrigerator knows—often insufficient time to make good energy-efficiency decisions;
- so-called “split incentives,” denoting cases in which decisions about energy efficiency often are made by those who do not pay the utility bills, and therefore will neither reap the benefits of improved efficiency nor bear the cost of poor efficiency;
- capital market failures, whereby customers may lack access to—or face competing demands for—the funds needed to make structural improvements or replace major pieces of equipment to improve efficiency;
- behavioral constraints such as inertia, limited attention, or heuristics that may lead consumers to make less than perfectly rational decisions; and
- knowledge spillovers that cannot be captured by the manufacturer and lead to underinvestment in new energy-efficiency technologies.
Potential policy solutions to these barriers are listed in Table 4-1 and discussed briefly in the subsections below. More detailed discussion of these solutions is provided in the section that follows on energy-efficiency policies in the United States.
Incorrect Energy Prices
If the externalities associated with electricity production and use are not reflected in their costs, incentives to invest in energy efficiency will be blunted. Since prices will not reflect the full societal cost of producing electricity, public intervention may be warranted to take prices to their correct levels. Potential solutions include a carbon price, other energy taxes, and carbon-trading instruments. Additional benefits of a price on pollution are addressed in Chapter 2.
|Incorrect energy prices||Carbon price
Other energy taxes
|Inadequate and imperfect information||
EnergyGuide/Energy Star™/Leadership in Energy and Environmental Design (LEED) labels
Comparative bills/home energy audits
|Capital market failures||
Third-party energy-service providers
On-bill repayment programs
EnergyGuide/Energy Star/LEED labels
Comparative bills/home energy audits
Appliance standards Building codes
Investments by the federal government
R&D tax credits
While a carbon price would help spur investments in energy efficiency, other barriers, discussed below, may inhibit the adoption of cost-effective energy-efficiency measures (Ryan et al., 2011). Moreover, the effectiveness of increased electricity prices in inducing conservation is limited by the very low measured price elasticity of demand for electricity, especially in the short term. Measures of the short-term price elasticity of demand for electricity range from 0.14 to 0.44 in absolute value for residential customers, meaning that a 10 percent rise in the price of electricity will reduce the quantity of electricity consumed by only 1.4 to 4.4 percent (Gillingham et al., 2009, Table 1). Over the longer term, these efforts would be more fruitful. The long-term price elasticity of demand for electricity is estimated to be between 0.32 and 1.89 (values greater than 1.0 mean that a 10 percent rise in the price of electricity would be expected to lead to a decline in electricity demand of more than 10 percent).
Finding 4-1: A more accurate price for electricity that includes its full social costs might help spur energy efficiency in the long run; however, higher electricity prices may be insufficient to overcome all market and behavioral failures that inhibit the adoption of energy-efficiency measures.
Another issue related to the price of electricity is that the bills of many consumers reflect the average cost of electricity, which may be lower than the incremental cost of producing electricity at peak times. Consumers then have less incentive to reduce their consumption during those peaks. Measures intended to move toward real-time pricing for electricity (or time-of-day pricing)—for example, by encouraging the adoption of advanced metering infrastructure or “smart meters”—could help reduce peaks in demand and perhaps overall electricity usage as well (Cappers et al., 2016).7 Because electricity produced during peak periods is usually generated from fossil fuels, reductions in peak demand could reduce pollution. In addition, evidence indicates that some form of time-of-day pricing would reduce overall demand for electricity.
Previous estimates of reductions in energy use induced by the adoption of smart meters range from 5.7 to 17 percent.8 The 2009 American Recovery and Reinvestment Act (ARRA) included $3.4 billion for the Smart Grid Investment Grant program, which promoted investments in smarter grid technologies, tools, and techniques, including funding for advanced metering infrastructure. Ten utilities were funded to undertake consumer behavior studies. A number of such
7 A full estimate of the impacts of time-varying rates on different consumers requires analysis of how changes in the aggregate load curve impact investment requirements and market prices.
8Houde and colleagues (2013) find a 5.7 percent drop in energy use only for the first month; Faruqui and colleagues (2010) find average energy use reduced by 7 percent; Gans and colleagues (2013) report that energy use in Ireland dropped by 11-17 percent.
studies will be conducted to investigate consumer acceptance, retention, and response to time-based rates (Cappers et al., 2016). The first consumer behavior study looked at differences in opt-in versus opt-out time-of-use metering programs, finding that many customers who defaulted to a time-of-use rate appeared to be better off, although some “inattentive” consumers who failed to opt out may have incurred higher electricity bills because of time-of-use pricing (Cappers et al., 2016).9
While utilities have reasons other than energy efficiency to deploy advanced metering infrastructure (addressed in more detail in Chapter 6), it is unclear whether the costs of deploying smart meters and the necessary communication infrastructure are great enough to dwarf the energy-efficiency savings to residential customers associated with these investments at this point in time (see Ramos et al., 2015 [citing Conchado and Linares, 2012]).
Inadequate and Imperfect Information and Split Incentives
Consumers do not directly observe the amount of electricity that is being used to wash clothes or dishes, keep a house at a certain temperature, or provide adequate lighting (see Ramos et al.  and Gillingham et al.  for more detailed analysis of such information failures). Additionally, it can be difficult for consumers to translate energy use into its cost, especially over a period of time, or to determine the savings derived from an energy-efficient device. Additionally, as discussed below, there may be cases in which the purchaser of an appliance is not the same as the person who uses it and pays the electricity bill, which may eliminate incentives for investments in energy efficiency.
A number of policies have been instituted to address the information failures leading to an energy-efficiency gap. These policies address energy use in buildings (for lighting, space heating, and cooling), as well as energy use by appliances. They include energy certificates (labeling such Energy Star™, EnergyGuide, or Leadership in Energy and Environmental Design [LEED] labels) and comparative bills/home energy audits whereby consumers receive personalized information on their electricity use compared with that of their neighbors and receive tips for reducing their electricity consumption (Allcott and Rogers, 2014).
As noted above, another informational failure relative to energy efficiency arises when the purchaser of the appliance is different from the user who is responsible for the electricity bill. Studies to date provide support that such split incentives exist (Davis, 2010; Gillingham et al., 2012). One example is set-top boxes that are owned by cable or telephone companies and are rented to
9 Suggesting that utilities could use focus groups or other forms of market research to make these customers aware of the transition to time-of-use pricing, better understand their options, and more easily navigate the opt-out process if they do not want to make the transition.
consumers. 10 Another is the landlord/tenant situation, where the landlord is responsible for the purchase of refrigerators, heat pumps, water heaters, and other appliances, but the electricity bill is paid by the tenant.11 In the United States, 37 percent of households are in rental housing, and the vast majority (86.4 percent) of those renters are responsible for paying their own electricity bill (Census Bureau, 2014). There is evidence of split incentives causing differences in behavior around energy efficiency: renters were found to be 1 to 10 percentage points less likely than owners to have Energy Star appliances (Davis, 2010). Potential policy solutions to the problem of split incentives include appliance standards and building codes.
Capital Market Failures
Energy-efficiency measures generally require a large up-front investment, which may be returned through a stream of smaller energy payments in the future. If some potential purchasers are unable to obtain credit, underinvestment in energy efficiency may result. As yet, there is little evidence regarding the size of this particular market failure (Gillingham et al., 2009).
Potential solutions to this barrier lie with third-party energy-service providers who pay the capital cost of an investment and receive a share of the resulting savings as payment, and on-bill repayment programs whereby the investments are recovered through charges on utility bills. Much more work is needed to identify and design appropriate policies for addressing failures to adopt energy-efficient technologies because of imperfect capital markets.
There are a variety of reasons why consumers, even when they have the correct information about energy prices and the net benefits of adopting an energy-efficient technology, may still choose not to adopt such technologies. For example, people put off decisions or find decision making difficult, or must weigh other characteristics (such as location, price, or size) in choosing a house (Allcott and Mullainathan, 2010; Gillingham et al., 2009). Consumers also may fear hidden costs or high transaction costs. In some cases, consumers may find qualitative attributes of the new technology less desirable than those of the existing technology (such as the inability to use a dimmer switch with compact fluorescent lighting) (Broderick, 2007). Hidden or high transaction costs that
10 The Federal Communication Commission (FCC) has issued a notice of proposed rulemaking (NPRM) that would allow consumers to purchase set-top boxes in the commercial market. See FCC NPRM MB Docket No. 16-42 and CS Docket No. 97-80, adopted February 18, 2016.
11 A comparable split can occur in commercial buildings if the specification and operation of the heating, ventilation, and air conditioning (HVAC) system are controlled by the building owner.
make adoption of an energy-efficient product or service more difficult also may affect consumers’ purchasing decisions. An example of these costs is found in weatherization assistance programs, which require residents to be available for at least two site visits (Fowlie et al., 2015). Consumers also may resist purchases that have high up-front costs, even when they recognize that the benefits in the long run are positive.
Another important factor is a fragmented retail market and a stovepiped energy policy landscape. Consumers cannot go to a single source to manage their energy requirements efficiently. Electricity and fuel supplies, distribution services, lighting, appliances, heating and cooling, building shell improvements, and control systems often are provided by different vendors, none of which necessarily have an incentive to optimize the consumer’s overall energy usage. The advent of advanced metering infrastructure, platform markets, data analytics, and “intelligent efficiency” providers that leverage increased data availability could help change the fragmented efficiency landscape. This is a potentially positive opportunity that is still in its early stage of development.
Any interventions designed to provide behavioral incentives on energy efficiency need to be scalable, to incorporate careful impact evaluation protocols that include control group comparisons and randomized field trials, and to have observable and measurable outcomes (Allcott and Mullainathan, 2010). One example is comparative bills or home energy audits, discussed above. A study of the effectiveness of these programs found that the monthly reports received by consumers provided cues that induced energy conservation, which persisted even after the program was terminated (albeit with some backsliding) (Allcott and Rogers, 2014). This finding suggests that further research on behavioral interventions could lead to positive economic benefits (Allcott and Mullainathan, 2010). Other potential policy solutions to behavioral constraints include EnergyGuide/Energy Star/LEED labels. In addition to residential and commercial customers, industrial customers may benefit from policies targeted to changing behavior (Gosnell et al., 2016). As with the problem of split incentives, moreover, appliance standards and building codes hold potential for addressing behavioral constraints.
As discussed in Chapter 3, electricity is an example of a general-purpose technology—an innovation technology that contributes to technological dynamism. Such technologies are subject to large knowledge spillovers—a form of market failure that prevents investors in innovation from realizing the full benefits resulting from their R&D investments. This problem can be addressed through government investments, tax credits for R&D, or inducement prizes (discussed in detail in Chapter 3).
The federal government can help overcome many of the obstacles to efficiency discussed above through indirect programs such as energy labeling, appliance standards, and building codes. The federal government also plays an important role in spurring innovation in energy efficiency through the R&D programs of the Department of Energy (DOE). In addition, it is poised to lead by example through direct efforts to promote energy efficiency in the 500,000 buildings it owns or operates. State and local governments also offer incentives to utilities to encourage retail customers to adopt energy-saving measures, among other incentives.
Energy Labeling and Certificates
Energy labels on appliances provide information about the energy savings that can be realized from adopting more energy-efficient appliances or equipment or assure consumers that a product is more efficient than the average appliance on the market. Energy labeling represents an inexpensive source of information on the operating costs of different appliances. While willingness to pay for labeled appliances varies across consumers, appliances, and states, evidence suggests that consumers may value the information provided by these labels (Ramos et al., 2015).13 Evidence also indicates that consumers may trust labeling by the government more than that by appliance manufacturers or other private parties (Banerjee and Solomon, 2003; but see GAO, 2010 [showing that Energy Star’s certification process is vulnerable to fraud and abuse]).
EnergyGuide labels, administered by the Federal Trade Commission (FTC), are mandatory energy usage labels.14 They apply to certain consumer products, such as clothes washers, refrigerators, freezers, televisions, water heaters, dishwashers, air conditioners, and boilers. EnergyGuide labels inform
12 In addition to the policies to spur energy-efficiency deployment described here, the Qualified Energy Efficiency investment tax credit, Internal Revenue Code (IRC) Section 25C, provides a 10 percent credit for the purchase of qualified energy-efficiency improvements to existing homes. The maximum credit for a taxpayer is $500, and no more than $200 of the credit can be attributed to exterior windows (in 2009 and 2010, the maximum credit was $1,500) (IRS, n.d.; NRC, 2013c). The credit is allowed for qualifying property in service through December 31, 2013. Additionally, IRC Section 25D allows for a credit for qualified expenditures made by a taxpayer for residential energy-efficient property placed in service before January 1, 2017.
13 However, see Newell and Siikamäki (2013) (showing that the impact of Energy Star certification may be due to a perceived endorsement of a model rather than the information provided) and Houde and Aldy (2014) (finding that net energy savings are small when a labeling program promotes products that have a high market share).
consumers of a product’s projected energy use, efficiency, and/or cost, based on DOE test procedures.
Energy Star, administered by the Environmental Protection Agency (EPA), is a voluntary label applied to more efficient heating, ventilation, and air conditioning (HVAC) equipment, lighting, home electronics, office equipment, and other appliances. The label is based on whether the appliance exceeds federal minimum efficiency standards by a certain percentage, which varies over time “depending on proportion of certified products offered on the market, the market shares, and the availability of new technologies” (Houde, 2014). In addition, certain state rebate programs for appliances are tied to Energy Star certification (Houde, 2014). Energy Star products are currently based on self-certification by manufacturers. In the absence of verification, products that do not meet the standard may end up on the market with the Energy Star label (GAO, 2010).
Although Energy Star was designed to supplement the EnergyGuide program, a consumer seeking to purchase a certain appliance that is covered by both programs would be confronted with various logos and labels containing different information using different formats. Thus the programs would be more effective if the relevant agencies attempted to harmonize their approaches and present the consumer with a common, consistent performance label wherever possible. Furthermore, while the Energy Star label indicates that a product meets a single, category-specific energy-efficiency benchmark, experience from other countries indicates that multitiered labeling by categories, such as the graded approach in the European Union, work well (CLASP, 2005).
Although some evidence suggests that too much information overwhelms consumers, other evidence indicates that better information leads to better outcomes. In one study, consumers were presented with information on how their state’s energy usage and prices were higher or lower than the national average used on EnergyGuide labels (Davis and Metcalf, 2014). Those consumers whose state’s energy usage and prices were higher than the national average were more likely to make a more energy-efficient purchase relative to consumers from states with lower energy prices and use.
While these energy labeling programs may be effective at reducing energy use in a cost-effective way, there is evidence that the state and utility energy rebates associated with Energy Star products may not be cost-effective (Alberini and Towe, 2015). Most consumers purchase a new appliance when their old one is no longer working, and it is unclear how many consumers would have purchased a more energy-efficient appliance even without the rebate (the “free rider” problem)15 (Alberini and Towe, 2015; Boomhower and Davis, 2014 Houde and Aldy, 2014; Malm, 1996 [finding that 73-92 percent of program participants were free riders]). While replacing appliances with more energy-
15 Free riders are program participants who would have participated without any intervention.
efficient versions does reduce energy use, it may be that energy-efficiency standards (discussed below) would be a cost-effective option (Alberini and Towe, 2015). Where products have a high market share, there is evidence that a labeling program (especially one combined with rebates) produces small net benefits because of the free rider problem (Houde and Aldy, 2014). In the case of a product with a high market share, one study found that 73 to 92 percent of program participants would have purchased the product without any incentives or labels (Houde and Aldy, 2014).
There are also energy labeling programs for construction of new buildings, including LEED16 and the Energy Star certification. Researchers have found that commercial buildings certified under these two programs in the United States have higher rents, higher selling prices, and higher occupancy rates relative to uncertified buildings (Ramos et al., 2015) [see Table 2 for a literature survey]). The literature on European buildings reports more varied results, with some researchers finding no effects of certification. Moreover, the actual energy efficiency of certified buildings remains subject to debate (compare Kahn et al.  and Newsham et al. ).
Research on the impact of energy labels on the residential sector has yielded mixed results (Ramos et al., 2015; Walls et al., 2016). While some studies have shown that homes with LEED or Energy Star certification do use less energy than noncertified homes, the reduction in energy use is not always reflected in their selling price (see, e.g., Walls et al., 2016; but see Ramos et al., 2015, p. S21 [stating that “the market has not yet been able to generate enough data to estimate the effect of introducing certificates on energy demand, neither at the aggregate nor at the disaggregate level”]). In addition, with both commercial and residential buildings, it is unclear whether certified buildings sell for higher prices because of the perceived energy savings, or they are perceived as having higher-quality building materials or better designs (Gillingham et al., 2009; Ramos et al., 2015). It may be that labels provide no additional information about net savings to the consumer, but a simple endorsement of a product may improve customer confidence in the product (Brounen and Kok, 2011; Newell and Siikamäki, 2013). In addition, certification at the state or local level may confound the perceived impact of LEED or Energy Star certification. Local certification may go beyond energy efficiency to include water efficiency, landscaping choices, and building materials, and the coexistence of such certifications makes it difficult to determine which of them is affecting consumers’ willingness to pay for a certified home (Walls et al., 2016).
16 LEED certification is administered by the U.S. Green Building Council.
DOE announces and implements minimum efficiency performance standards (MEPS)17 for a variety of residential appliances, including central air conditioners and heat pumps, clothes washers and dryers, major kitchen appliances, and room air conditioners (Office of Energy Efficiency and Renewable Energy [EERE], DOE, current rulemakings and notices). The standards do not cover some classes of appliances, such as computer and battery backup systems, portable ovens, portable air conditioners, set-top boxes, and televisions (although some states have mandatory standards for these appliances). According to EERE, these standards apply to more than 50 categories of products that in the aggregate cover about 90 percent of home energy use, about 60 percent of commercial building energy use, and almost 30 percent of industrial energy use (DOE, n.d.-b).
MEPS work to remove the least efficient appliances from the marketplace, and most researchers agree that these standards appear to have improved consumer welfare (Houde and Spurlock, 2015; Taylor et al., 2015; but see Gayer and Viscusi, 2013 [arguing that consumer welfare is increased only if consumers are not behaving rationally]). The energy efficiency of many appliances covered by these standards has increased substantially, and many consumers are choosing to buy products that exceed the standards (Taylor et al., 2015). The result has been an average energy efficiency of purchased products that exceeds the MEPS requirements (Taylor et al., 2015). While many of the products have not seen price increases, unregulated quality dimensions (e.g., performance, capacity, noise) and product diversity have improved even as the standards have become more stringent (Houde and Spurlock, 2015 [citing Dale et al., 2009; Spurlock, 2013; Allcott and Taubinsky, 2015, to show mixed results on price impacts]). Reliability also has not been harmed by appliance standards; the rate of significant repairs over 5 years of product ownership generally declined from the time appliances were first subject to federal MEPS (Taylor et al., 2015).
These standards have led to much smaller increases in appliance prices than expected ex ante using engineering models, and in fact led to only modest increases (Dale et al., 2009; Houde and Spurlock, 2015; Taylor et al., 2015). Refrigerators are a widely cited example of an appliance whose energy efficiency rose simultaneously with declines in prices. Figure 4-1 shows the
17 In the 1970s and 1980s, appliance standards were used in states such as California, New York, and Florida. A federal program that included energy targets was established in the Energy Policy and Conservation Act of 1975, although the federal minimum standards did not preempt state-level standards until the passage of the National Appliance Energy Conservation Act of 1987. New categories for federal standards were added in the Energy Policy Act of 1992, the Energy Policy Act of 2005, and the Energy Independence and Security Act of 2007. The Energy Independence and Security Act of 2007 also required that DOE maintain a schedule for regularly reviewing and revising all standards (DOE, n.d.-b).
decline in the energy used by a new refrigerator from the mid-1970s through 2008 along with the enactment of energy efficiency standards. Simultaneous increases in product quality and diversity together with no or little change in prices may at first appear counterintuitive. It may be, however, that stringent standards led manufacturers to compete on other quality dimensions because they all had to meet a certain minimum efficiency standard (Houde and Spurlock, 2015).
One unanswered research question is how great an impact appliance standards have on inducing technological innovation to meet the standards. It is likely that standards did spur manufacturers to innovate to ensure that their products would exceed the minimum standards, although how much of that innovation was due to the standards, to increasing energy prices, or to exogenous R&D efforts is unclear (see Newell et al., 1999).18
Legal requirements mandate that the standards be reviewed at least every 6 years,19 and test procedures must be reviewed at least every 7 years to determine whether updates are warranted.20 DOE also evaluates new product categories for standards as opportunities for energy efficiency emerge. Other countries use a different type of appliance standard. The Top Runner approach used in Japan is somewhat analogous to the CAFE (Corporate Average Fuel Economy) standards for automobiles in the United States in that it uses a market-weighted average21 rather than a minimum allowable efficiency. Top
19 United States Code, 42 U.S.C. 6295 (m).
20 United States Code, 42 U.S.C. 6293 (b).
21 A market-weighted average is calculated for each machinery or equipment category for each of the companies that manufacture (or import) machinery or equipment covered by
Runner sets a high efficiency level for a future date, and the market must meet that weighted-average efficiency.
Efficiency standards and policies can be made more stringent, which may spur technical innovation and market competitiveness. Furthermore, appliance standards may help address the problem of split incentives since a landlord has no choice but to provide appliances with better energy efficiency (Gillingham et al., 2012). More research is needed on the connections and interactions between efficiency standards and other policies designed to spur innovation in energy efficiency, especially any interaction between standards and policies to internalize prices on pollution. 22
Recommendation 4-1: DOE should on an ongoing basis set new standards for home appliances and commercial equipment at the maximum levels that are technologically feasible and economically justified.
Given the increasing share of electricity going to television and electronics, DOE might consider expanding appliance standards to include these “nontraditional” appliances that are not subject to MEPS. The committee also recognizes that energy efficiency is not a societal value that trumps all other values. Indeed, in promulgating these standards, cost-effectiveness to consumers, economic impacts on manufacturers, impacts on product performance, impacts on competition, and other factors explicitly cited in the law need to be taken into account. 23
Building Codes and Retrofits
Building codes (and other efforts to improve new construction) can have a positive effect on improving overall energy efficiency. Building codes are the primary policy instrument for influencing the energy efficiency of newly built or renovated buildings and can provide information on best practices to the multiplicity of builders and contractors across the United States. National model energy codes exist (see below), but building codes are determined by state and local governments.24
the Top Runner program. This average is based on the volume of each of the products shipped and their efficiency (METI Agency for Natural Resources and Energy, 2010, p. 26).
22 In the case of sulfur dioxide, research suggests that direct measures for reducing the pollutant encouraged only cost-reducing innovations, while economic incentives such as pricing the pollutant encouraged both cost-reducing and emission-reducing innovations (Harrington and Morgenstern, 2004).
23 The Energy Policy and Conservation Act of 1975, Section 325(o)(2)(B).
As noted earlier, energy-efficiency measures and migration to states that are warmer in the winter both have contributed to a decline in residential energy use. Despite a large increase in the number of electronic devices and appliances used in households, there has been only a small increase in overall residential electricity demand (EIA, 2013b). According to the most recent survey, newer homes (built between 2000 and 2009) used only 37.1 million Btu (MMBtu)/ft2 in 2009 for space heating and cooling, compared with 51.6 MMBtu/ft2 for homes built before 1940 (EIA, 2009b, Table CE2.1).
With respect to rental properties, because builders or property owners generally choose the windows, amount of insulation, and other energy-efficient aspects of the physical property, the electricity payer may have few choices for investing in energy efficiency—the phenomenon of split incentives discussed earlier. While landlords may be able to capture some of the value of energy efficiency through higher rents, renters may value other housing characteristics, such as location, size, number of bathrooms, and other factors, more than energy efficiency, even if their electricity bill is higher in a less energy-efficient rental unit. One study found owner-occupied homes to be 12 to 20 percent more likely to have insulation than rental units, even after controlling for observable characteristics of property, occupant, and neighborhood (Gillingham et al., 2012).
Although evidence shows that energy use for heating and cooling buildings has decreased over time, it is unclear how much of that decline is due to building codes versus improvements in building materials that would have occurred in the absence of the codes. Research on the link between building codes and improved efficiency is mixed, and more work is needed to determine whether such programs are cost-effective (Horowitz and Haeri, 1990; Jacobsen and Kotchen, 2013; Kotchen, 2016; Levinson, 2014b).
The national model energy codes—the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE)25 code and the Interventional Energy Conservation Code® (IECC)—have historically addressed only construction and renovation; there are no requirements covering ongoing operations, such as that energy use be measured; that building operators be trained to operate energy systems properly; or that energy systems be regularly tuned, as is the case for automobiles. Buildings seldom operate the way they were designed to operate, and even if commissioned properly, seldom continue to perform optimally. Lack of controls or inability to use existing building automation properly may be leading to a substantial increase in energy use (Mills, 2011; Piette et al., 2012). Sensors and communication and computational capabilities can help unlock a substantial portion of these operational savings.
25 ASHRAE was formed by the 1959 merger of the American Society of Heating and Air-Conditioning Engineers, founded in 1894, and the American Society of Refrigerating Engineers, founded in 1904 (https://www.ashrae.org/about-ashrae; International Energy Conservation Code, http://www.iccsafe.org/codes-tech-support/codes/2015-i-codes/iecc). The IECC was created by the International Code Council in 2000.
To enable enhanced energy performance in commercial buildings, the key is to install sensors that measure building energy consumption, to link those sensors to a microprocessor, and to apply software to analyze the data and provide insight into opportunities for energy savings. Advanced data analytics supported by advanced or interval meter infrastructure also can assist significantly. For example, IBM’s sustainability efforts through its Smarter Planet platform rely on such an information technology-based approach (Hudson, 2012).
Recommendation 4-2: DOE’s Office of Building Technologies should continue to partner with ASHRAE and the developers of the IECC in the development of these model codes, and should support them in efforts to compile best practices for improving the energy efficiency and operation of building systems.
Attempts to reduce energy use in residential dwellings need to address not just new construction and renovations—the target of building codes—but also retrofits to the existing housing stock. Houses have a long life—residential buildings have a typical service life of 60 to 100 years (Jaffe et al., 2004)—and building standards for new construction have no effect on the existing housing stock. However, there is little evidence that retrofits have improved energy efficiency. Studies of energy-efficiency retrofits during the 1970s and 1980s have found that the actual energy savings achieved were far below the engineering estimates (see Geller and Attali  for a review of the literature). More recently, a study of the weatherization assistance program in Michigan found that, even accounting for the social benefit of reduced energy use, the rate of return on weatherization was negative and that the up-front costs were twice as great as the energy savings (Fowlie et al., 2015).
Better information on actual energy use is needed before any assessment of a government energy-efficiency program can be carried out. One way to achieve that goal is to improve engineering models so they more closely resemble actual use. Energy-efficiency policies could also be improved by field studies, conducted by both DOE and the state public utility commissions, relying on state-of-the-art evaluation approaches. DOE has already solicited ideas for improving the accuracy of its audits, issuing a request for information (RFI) on means of improving savings prediction methods for residential energy-efficiency upgrades. DOE expressed its interest in information on the current state of the art of savings prediction methods, forthcoming advances that could improve the accuracy and/or reduce the costs of these methods, and the potential market implications of improved methods. In addition, DOE asked for information on “other factors that may influence the accuracy of modelled predictions such as improved field guidance, technician training and certification, or benchmarked models to a nationally accepted DOE prediction model such as EnergyPlus.”
A home energy audit, also known as a home energy assessment, is one way for a homeowner to obtain personalized information on home energy consumption and measures that can be taken to reduce it (Ramos et al., 2015). Costs for home energy audits can range from $300 to $500 (Palmer et al., 2013 [average cost is $349]26); however, some local governments and utilities provide subsidies to pay for these audits (Alberini and Towe, 2015 [analyzing free energy audits in Maryland]).
Despite the potential for bridging informational and behavioral failures, research on the cost-effectiveness of energy audits is limited (Alberini and Towe, 2015; Frondel and Vance, 2012; Palmer et al., 2015; Ramos et al., 2015). A study of free audits in Maryland found that residential energy audits reduce energy usage by about 5 percent, with the cost per ton of carbon dioxide emissions abated ranging from just under $50/ton to nearly $70/ton (Alberini and Towe, 2015).27 Even when homeowners pay for audits, they may not follow through on the resulting information (Palmer et al., 2015). Moreover, only about 4 percent of U.S. homeowners have had an energy audit (Palmer and Walls, 2015). Thus while these audits offer the promise of reducing energy use, additional research is needed to ensure that this promise is realized and to illuminate how information policies can be made more effective.
The Role of Innovation
Chapter 3 provides a detailed discussion of innovation, and notes the potential underinvestment in R&D because of knowledge spillovers. Federal funding for R&D for energy conservation and end uses is quite small, amounting to less than $1 billion in fiscal year (FY) 2013 and accounting for about 20 percent of federal spending on energy conservation and end uses when energy assistance programs are excluded (EIA, 2015c, Table 1).28 As discussed in Chapter 3, without government support for energy innovation, unpriced environmental externalities may reduce incentives for private-sector innovation in new technologies, including those that reduce energy use (Gillingham et al., 2009 [citing Goulder and Schneider, 1999; Jaffe et al., 2005; Schneider and Goulder, 1997).
One important research question has to do with the link between energy-efficiency policies, such as energy labels or standards, and technological change. In other words, do energy-efficiency policies induce greater investments in technological change to improve energy efficiency? While more research on this
27 The authors caution about selection bias in comparing energy use between homeowners who undergo an audit and those who choose not to do so.
28 This excludes direct expenditures on the Low-Income Home Energy Assistance Program, which assists low-income households with their energy bills.
question is needed, preliminary findings show that a country’s commitment to increasing energy efficiency, along with funding for energy-efficiency research, does lead to a higher probability of innovation (Verdolini and Galeotti, 2011). Other researchers have found that increased regulatory stringency leads to more spending on R&D; however, no relationship has been found between increased costs to comply with more stringent requirements and innovation (as measured by successful patent applications) (Jaffe and Palmer, 1997).
One example of DOE’s R&D program is a joint effort with industry that has been ongoing for more than a decade according to a road map, focused on the development and implementation of improved light-emitting diodes (LEDs). The program has funded fundamental science, core technology research, product development, manufacturing R&D, and commercialization support, and has included prizes and test procedures and standards. The program has resulted in LEDs that use 84 percent less energy than standard incandescent bulbs while giving off the same amount of light and lasting 25 times longer (25,000 hours versus 1,000 hours). According to EIA, the efficiency of LED bulbs (light output per unit of energy consumed) has risen even as the costs per bulb have fallen dramatically since 2010 (from nearly $70/bulb in 2010 to $10/bulb in 2014), and costs are predicted to continue to decline through 2020 (EIA, 2014g).
Other examples of DOE R&D include a number of programs supported by the Advanced Research Projects Agency-Energy (ARPA-E), including DELTA (Delivering Efficient Local Thermal Amenities), aimed at reducing the costs of heating and cooling buildings through the development of technologies such as on-body wearable devices and more options for maintaining occupants’ comfort within a building; BEETIT (Building Energy Efficiency Through Innovative Thermodevices), focused on enhancing energy efficiency, reducing greenhouse gas emissions, and reducing consumer costs for cooling commercial buildings; and SWITCHES (Strategies for Wide-Bandgap, Inexpensive Transistors for Controlling High-Efficiency Systems), aimed at developing next-generation power switches that could increase the efficiency of appliances and lighting.29 More recently, ARPA-E announced a new program to develop innovative window coats and window panes that could improve the efficiency of existing single-pane windows (ARPA-E, 2016).
In addition to research on technologies for improving energy efficiency, DOE’s Quadrennial Technology Review acknowledges the need for further behavioral research, noting that estimates of the energy-savings potential from research on consumers’ decision making in the residential building sector are substantial (DOE, 2015b, Section 10.2.4).
R&D drives a much longer-term process that ultimately brings much-improved energy-efficiency technologies to the market. Accordingly, R&D needs to focus on both basic and market-oriented research.
Recommendation 4-3: DOE should increase its investments in innovative energy-efficiency technologies; improve its ability to forecast energy savings from these technologies; and, in conjunction with other agencies, obtain data with which to develop behavioral interventions for improving energy efficiency.
Energy-Efficiency Technologies in the Industrial Sector
The industrial sector accounts for just under a third of the energy use and carbon dioxide emissions30 in the United States (EIA, 2016a, Table 2) (and about 54 percent globally [EIA, 2016b]). Of course, the energy efficiency of industrial buildings can be increased through many of the same approaches for commercial buildings discussed earlier in this chapter. More important, changes in industrial processes have reduced energy use in the past (see, e.g., Levinson, 2015 [showing that reductions in emissions in the manufacturing sector are driven by changes in technique rather than industry composition]) and offer significant promise for improving energy efficiency in the future. However, there are barriers that may be inhibiting those investments. New equipment is considered an asset and is subject to different accounting treatment from that applied to energy bills or other operations and maintenance (O&M) expenses. Thus managers may be reluctant to make large investments in new equipment if the payoff of reduced O&M expenses is credited to another part of the company. Uncertainty about future energy prices is an additional barrier to investments in industrial energy efficiency. Policies to reduce that uncertainty—including uncertainty about a price on carbon—could increase action by industry to reduce energy use.
In addition to improvements in industrial buildings, more energy-efficient production processes could help reduce electricity consumption. Efforts to improve information on energy use or change the behavior of workers to improve energy efficiency could prove valuable in the industrial sector as well as in the residential and commercial sectors (see, e.g., Gosnell et al., 2016 [showing that performance information, personal targets, and prosocial incentives induced pilots to improve their fuel efficiency on flights]).
The efficiency of industrial processes has improved, leading to much lower energy use in this sector, due partly to strategic energy management programs and partly to sectoral changes. Motor systems (e.g., pumps, fans, air compressors, and motor-driven industrial processes) can be improved by sizing correctly to load, using speed control, improving maintenance, enhancing the efficiency of underlying motor drives and associated components, and other
30 Assuming that carbon dioxide emissions are proportional to energy use.
measures. Equipment standards for motor drives, along with guidance on the other efficiency measures, can help boost motor system efficiency significantly.
Another potential energy-efficiency measure is for industrial customers to produce heat and power simultaneously from the energy they have at their disposal; this cogeneration is known as combined heat and power (CHP). Waste heat from electrical processes or from the generation of electricity on site is recycled to produce additional electricity and steam that can be used to warm buildings or assist in industrial processes. According to the most recent survey, fewer than 10 percent of industrial establishments use cogeneration technology (EIA, 2010, Table 8.2).31 While approximately 85 gigawatts (GW) of CHP is installed in the United States, an additional 130 GW could be produced by the commercial/institutional and industrial sectors (DOE and EPA, 2012). The technical potential of CHP could rise further to about 200 GW if the industrial sector sized its CHP systems to sell power, although there are a number of market and policy barriers to realizing this potential (DOE and EPA, 2012).
Encouraging companies to institute a management structure focused on energy savings—strategic energy management—also could help realize the potential of operational/behavioral energy efficiency. In addition, encouraging equipment-based energy efficiency on an ongoing basis could help incentivize companies to invest in cost-effective energy-efficiency measures.
Direct Federal Government Efforts to Promote Energy Efficiency
Positive spillovers, such as reducing the cost of production by producing more or learning by doing, may occur in energy efficiency. In addition, there may be a “free driver” effect whereby the first customer provides information to later customers about the quality of a new energy-efficient technology or appliance. These effects may justify direct federal government efforts to adopt energy-efficient technologies, although more research is needed to determine the magnitude of these effects.
The federal government owns or operates more than 500,000 buildings across the country, comprising more than 3 billion square feet of total floor space and accounting for 2.2 percent of all building energy consumption in the United States (DOE, 2011a). State and local government buildings account for nearly 10 percent of the U.S. total. The federal government spends roughly $7 billion each year on energy to heat, cool, light, and power federal buildings (WBDG, n.d.).32 State and local governments spend another $30-$40 billion per year on building-related energy consumption. Improving the energy efficiency of this building stock could reduce spending on electricity without impairing the services government currently provides.
32 Note that the federal government spends $20 billion annually on energy, but a large fraction of that amount goes to nonbuilding energy use.
The Federal Energy Management Program (FEMP) provides expertise, training, and other services to help federal agencies improve energy efficiency and increase the use of renewable energy (Sissine, 2015). It pursues those goals through project financing, technical guidance and assistance, and planning and evaluation. Last year, the White House issued an executive order on planning for federal sustainability in the next decade, which ordered the head of each federal agency to promote building energy conservation, efficiency, and management (White House, 2015). The federal government can continue to lead efforts to promote energy efficiency by doing the following:
- Continuing to lead in developing procurement practices for appliances and equipment that take life-cycle cost into account. Federal agencies are required to purchase Energy Star or FEMP-designated appliances and equipment where those products are available (FEMP, n.d.). However, the incremental benefits of using life-cycle costing versus the additional administrative burden need to be assessed. A previous National Research Council report on solid-state lighting includes the recommendation that the Office of Management and Budget develop criteria for determining life-cycle costs and including social costs in evaluations of energy purchases, and for incorporating this methodology into agency procurements (NRC, 2013b). The President’s Council of Advisors on Science and Technology made a similar recommendation (PCAST, 2010). Applying life-cycle costing ought not to be highly challenging; DOE has extensive expertise in this regard through the National Appliance Energy Conservation Act (NAECA) appliance and equipment energy-efficiency standards program (DOE, 2014b), which could serve as a template for other federal efforts.
- Evaluating the benefits of improving the energy efficiency of the Department of Housing and Urban Development’s (HUD) 1.2 million units of public housing. Existing accounting and procurement rules make it difficult to improve the energy efficiency of public housing because energy savings cannot be credited against capital expenditures. These rules need to be restructured. In addition, much of HUD’s building stock consists of large housing blocks built in the 1950s through the 1970s that are overdue for major capital upgrades. Similar blocks of midcentury social housing in Europe have been rewrapped with highly insulating facades and retrofitted with efficient equipment to achieve energy savings in excess of 50 percent. Additional benefits of such measures include increased resilience, reduced costs of maintaining the building envelopes, and reduced occupant health issues due to drafts and condensation. HUD could partner with local housing authorities and others to pilot rewrappings of midcentury housing blocks, including the development of scalable
financing models based on life-cycle assessments of costs and savings. To the extent that existing rules inhibit the energy efficiency of housing on military bases, those rules could be restructured accordingly.
- Taking the lead on contracting for services that gives third parties incentives to invest in energy efficiency. Various statutes and regulations authorize federal agencies to enter into contracts for their utility services, including electricity, hot water, and steam. The General Services Administration (GSA) is typically the lead federal contracting agency. Performance contracts such as energy savings performance contracts (ESPCs) or utility energy service contracts (UESCs), currently in use in the Department of Defense and other parts of the government, could serve as a model for the federal government’s providing incentives for the private sector to invest in energy efficiency, with the cost of the investment paid off through reduced energy or O&M costs achieved over the life of the contract.
Current budgetary rules may inhibit the federal government from fully investing in energy-efficiency technologies. A typical ESPC provides payment from the government to the vendor until the vendor’s costs have been covered and the contract expires. During the contract period, the government retains only a small share of the savings; after the contract expires, the government keeps all savings from the energy-efficiency investment. The Congressional Budget Office has identified two issues that may inhibit the government from engaging in these investments, even if they pay off over time (CBO, 2015). First, the additional spending for an ESPC is considered mandatory spending, while the potential future savings from the contract are considered discretionary spending. Second, while much of the additional spending for an ESPC falls within the 10-year cost estimate used by the Congressional Budget Office, much of the savings occurs over a much longer period of time.
Recommendation 4-4: The federal government should lead by example and ensure that federal facilities, including leased space, capture all cost-effective energy-efficiency opportunities, such as setting and achieving energy-efficiency targets, using performance contracting to achieve energy-efficiency gains, and procuring energy-efficient appliances and equipment.
State and Local Efforts to Improve Energy Efficiency
In addition to actions taken at the federal level, states are adopting important energy-efficiency measures. Because the retail market for electricity is
regulated at the state level (for investor-owned utilities), states can play an important role in providing appropriate incentives for utilities to adopt energy-efficiency measures, including utility and public benefit programs and policies (ACEEE, 2012). In addition, states provide building energy codes, policies encouraging CHP systems, and state government-led initiatives around energy efficiency. Twenty-four states have an energy-efficiency resource standard (EERS) of some kind; 16 of these states have achieved electricity sales reductions of nearly 1 percent or greater, while 6 have achieved savings of greater than 1.5 percent, on an annual basis (ACEEE, 2012). While higher electricity prices do influence and can assist in the uptake of energy-efficiency measures, the evidence suggests that state policies play a similar role (ACEEE, 2012; EIA, 2014f, Table 5.6.A).
Some energy-efficiency programs—such as advanced metering infrastructure and comparative energy bills—may be run more efficiently through utilities because of access to information or lower transaction costs. However, state programs designed to encourage reductions in electricity consumption that are led by the utilities (demand-side management [DSM] and efficiency programs) are hampered by the fact that many utilities recover their fixed costs through volumetric rates that are charged on a per kWh basis. Because revenues are tied to the volume of electricity sold, utilities have little incentive to take measures that would reduce consumption. Allowing utilities to recover their fixed costs through a fixed charge is one way to remove disincentives for utilities to invest in DSM programs, although there are also other mechanisms that effectively decouple revenues from rates to remove the disincentive for utilities to invest in energy efficiency (see, e.g., Arimura et al., 2011).33 Spending on DSM programs by utilities peaked in 1993, with combined customer incentives and other costs of about $4.5 billion in 2014 for all customer classes (EIA, 2014e, Table 10.6).34 The cost-effectiveness of DSM programs is still being debated, with savings ranging from 1 cent to more than 20 cents per kWh saved (Arimura et al., 2011 [citing Gillingham et al., 2006, for a literature survey].
Cities and local governments are also beginning to play leadership roles in deploying energy-efficiency measures. City leaders are looking for ways to reduce the substantial amount of energy waste in local buildings as one step toward making their communities more resilient and sustainable.
One caveat to keep in mind is that the demand for energy efficiency varies across states, in large part because of differences in electricity prices and usage. Great variation also exists within states as a result of heterogeneity among consumers due to differences in preferences, incomes, discount rates, and other factors. Small changes, such as a small change in interest rates, may make a
33 Additional programs include performance incentive programs and lost revenue recovery mechanisms.
difference between a consumer’s buying a more expensive but more energy-efficient appliance versus a less expensive, less energy-efficient appliance.
Energy-efficiency measures targeting electricity usage appear to have contributed to improvements in energy productivity during the past 30 years, likely saving many tens of billions of dollars. In particular, the committee considered evidence for an energy-efficiency gap—the difference between projected savings from avoided energy use due to energy-efficiency measures and the actual savings realized. To the extent that such a gap exists, there may be several barriers to higher utilization of energy-efficiency measures.
First, electricity prices do not, for the most part, incorporate the costs of pollution. Second, even if prices are corrected to include the costs of pollution, other market imperfections may limit consumers’ purchases. Information about energy use and price is not always readily available to consumers, and even when it is, consumers may be unable to translate price information into actual costs (or into actual savings in the case of energy efficiency). Additionally, consumers may be reluctant to make new purchases out of inertia, limited attention or heuristics. Moreover, the effectiveness of increased electricity prices in inducing conservation is limited by the very low measured price elasticity of demand for electricity, especially in the short term.
The committee notes that energy efficiency is also an area in which innovation is critical. DOE has several R&D programs under way. The solid-state (LED) lighting program is a joint program with industrial partners that has utilized a road map and produced sizable increases in performance and cost declines. ARPA-E has one program aimed at reducing the costs of heating and cooling buildings and another focused on developing next-generation power switches that could increase the efficiency of appliances and lighting. DOE should be encouraged and supported to create more R&D programs that can ultimately bring much-improved energy-efficiency technologies to the market.
Recommendation 4-5: The federal government, state and local governments, and the private sector should take steps to remove barriers to, provide targeted support for, and place a high priority on the development and deployment of all cost-effective energy-efficiency measures.