Energy Information Administration Home Page Welcome to the Energy Information Administration's Energy Efficiency Web Site.  If you need assistance in viewing this page, please call (202) 586-8800

  Home > Energy Users > Energy Efficiency Page > U.S. Energy Usage and Efficiency

United States Energy Usage and Efficiency:
Measuring Changes Over Time

BATTLES, Stephanie J. and BURNS, Eugene M.


Washington, D.C., United States of America

(17th Congress of the World Energy Council, Houston Texas, September 14, 1998).


    Increasing emphasis has been placed on energy efficiency as a vital component of the United States’ energy strategy. This was evident with the passing of the Energy Policy Act of 1992 (EPACT) [1]. EPACT promotes energy-efficiency programs such as building energy-efficiency standards, residential energy-efficiency ratings, and energy-efficient mortgages. It also encourages investments in conservation and energy efficiency for each of the sectors in the United States (U.S.).

    EPACT also includes provisions to obtain energy-efficiency information "with the objective of significantly improving the ability to evaluate the effectiveness of the Nation’s energy efficiency policies and programs."

    The U.S. Department of Energy (DOE), universities, trade associations, individual researchers, and others have dealt with efficiency assessment by publishing a variety of special analyses of energy efficiency. Detailed analyses have tended to cover limited sectors of the economy, while analyses with comprehensive scope have ordinarily used very broad-brush efficiency indicators that have limited explanatory power, and may be misleading. Most important, these products do not attempt to give a single comprehensive answer to the question "How is energy efficiency changing in the U.S. economy?"

    The concept of efficiency improvement is easy to rally behind as a general principle. Defining it, measuring it, and devising specific programs to encourage it are much more difficult tasks. In 1993, the Energy Information Administration (EIA) started a project to determine the most useful efficiency measures, the most appropriate data sources, and a workable methodology for combining separate sectors into a unified whole. Included in the project would be a historical database of energy use and energy-efficiency indicators, an assessment of how the indicators are changing over time, and measurements of how those efficiency changes have contributed to the trends in total U.S. energy demand.

    This paper will begin by presenting EIA’s history and role as an energy data source. Next presented will be a brief synopsis of the developments in EIA’s energy-efficiency project since it began in 1993. The following three sections present discussions using data that will be included in the Internet database. First will be a profile of energy use in the United States and a look at how this usage has changed between 1970 and 1995. This will be followed by a discussion of energy efficiency--first as to its definition and then its measurement. Two examples of energy-efficiency indicators--national and manufacturing sector indicators will be included in this discussion. International competitiveness makes energy use a vital concern for some manufacturing establishments. The third category of indicators, carbon emissions indicators, will be part of a discussion of energy use and indicators in the transportation sector. This sector is important because energy efficiency, the environment, and energy security are so intertwined. The paper will conclude with a summary discussion of why EIA’s energy-efficiency project has taken on new importance.

       Energy Information Administration: Who We Are and What Do We Do

    In 1977, the EIA was created as an independent statistical and analytical agency within the U.S. Department of Energy. The EIA's mandate in the newly-formed Department was to collect, analyze, and disseminate impartial, comprehensive data about energy--how much is produced, who uses it, and the purposes for which it is used. EIA's duties include collection of energy-related data from energy suppliers and energy consumers, tabulation and analysis of energy data for publication, and energy demand forecasting. In recent years, EIA has also produced estimates for U. S. emissions of greenhouse gases.

    As the agency charged with statistical and analytical responsibilities for U. S. energy data, EIA has both the data and the expertise to conduct studies of energy-efficiency measurement. EIA has access to data from a variety of supplier and consumer surveys. Furthermore, EIA's staff has the experience needed to make optimal use of the available data to address the complex issues of efficiency measurement. Finally, EIA's reputation for providing unbiased results, and for transparency of methodology, is important in dealing with energy efficiency and global climate change issues.

        Energy-Efficiency Project

         Initial Development Plan

EIA has taken a slow, deliberate approach toward defining energy efficiency and developing robust energy-efficiency indicators. The early phase included three major steps: (1) a methodological report, (2) a series of five workshops, and (3) the use of EIA’s web site.

In October 1995, EIA published the methodological report, Measuring Energy Efficiency in the United States’ Economy: A Beginning [2]. This report was called a "beginning" because it was not intended as a definitive statement on these issues, but rather, as a means of focusing the thinking of our customers and obtaining their ideas. Even in the writing of the report, DOE, EIA, and outside sector specialists furnished feedback in the form of written reviews. Draft chapters of the report were presented for comment at seminars as soon as they were available.

Early in 1996, EIA held five all-day workshops covering each sector discussed in the report: residential, commercial, transportation, industrial, and the economy as a whole. The focus of the workshops was the report. The five one-day workshops were held in Washington, D.C., in February, March and April of 1996. Over 120 experts, representing federal agencies, industry, associations, academia, and energy-efficiency advocacy groups attended the workshops. The participants, including engineers, economists, planners, and policy analysts, represented broad concerns as well as varied experiences in the analysis of energy efficiency. These workshops allowed EIA to obtain critical reviews from leading energy-efficiency researchers.

As another step in the project, EIA decided to place the report, the workshop comments, and a conversation area on the Internet. We used the EIA home page address as the site. A notice was sent to the workshop participants and other interested parties. Although we received several inquiries, comments, and questions about the initial report through our web site, the conversation area at the web site was not successful. From those who did respond, it seemed that they were expecting either a "true" listserver or a "chat" group.

          Latest Development Plans

    During the summer of 1997 the main task was to explore ways of incorporating as many as possible of the comments received from the report, workshops, and web site. The development of new indicators is now underway, and we plan to use the Internet as the main method of communication. We believe that the electronic communication will reach the maximum number of customers while conserving our resources.

    Under development is a database consisting mainly of energy use, energy efficiency, and carbon-emission indicators. Initially, only a modest database will be placed on the web site along with 1 to 2 page analyses on special topics. Additions will be made when further highlights or other more complex and sector-specific indicators become available. A query system is being developed whereby customers will be able to easily obtain data from the main tables in the database. As an additional method of communication, we will print booklets and brochures presenting highlights of the database and the most important findings.

        Energy Use in the United States

           United States Energy Use Relative to World Use

I n 1995, the U.S., with a population of 263 million people, used an estimated 95,300 PJ of primary energy. This estimate represents 25 percent of the world energy use while the U.S. has only 5 percent of the world population. By comparison, Japan--ranked 4th in energy use--used 6 percent of the world energy while having 2 percent of the world population [3].

Many factors affect the quantity of primary energy used by a country. Primary energy includes not only the energy directly consumed by the end-users of the energy, but also the losses associated with the generation and transmission of electricity. Thus, countries that are electricity-intensive will tend to have large primary energy requirements. In 1995, the U.S. used approximately 27 percent of the world’s electricity versus Japan’s 7 percent.

The level of economic production also affects the level of energy use. In 1995 the U.S. Gross Domestic Product (GDP) at market exchange rates was the highest of any country-- 5.5 trillion dollars (1987 dollars). The next largest economy, Japan, had a GDP of 3.0 trillion dollars (1987 dollars). Land area is also directly related to the level of transportation demand. For example, while the U.S. ranks 3rd in population and Japan ranks 9th, the population density for Japan (318/km2) is more than 10 times that of the U.S. (29/km2). The U.S. has more than three times as many motor vehicles as Japan [4].

        Energy Use in the United States: a Profile

In 1995 fossil fuels (petroleum, natural gas, and coal) accounted for 85 percent of the primary energy consumed in the U.S. Petroleum accounted for 38 percent of the energy used in the U.S. (Figure 1a). This was equivalent to 18 million barrels per day, with 12 million barrels per going to the transportation sector, mostly for motor gasoline [5]. Natural gas accounted for 25 percent (23,400 PJ) of the energy used in the U.S. Natural gas was used mainly in the industrial sector (45 percent of the natural gas used), followed by the residential sector (22 percent), and then the commercial sector (14 percent). Coal, accounting for 22 percent of energy use (20,800 PJ), is heavily used by the electric utility industry. The electric utility industry consumed 86 percent of the coal used in the U.S. In 1995, electric utilities consumed 33,416 PJ of energy, of which 54 percent was coal [6].

Electricity can be generated using conventional energy sources such as coal, but also from other sources such as nuclear energy and hydropower. This type of generated electricity accounts for 15 percent of energy use in the U.S. To measure the full effect of electricity use in the major end-use sectors (industrial, residential, commercial, and transportation), we need to take into account the losses incurred in the generation and transmission and allocate these losses to each of the sectors. Taking losses into account, in 1995 electricity use in the U.S. represented almost as much energy content as did petroleum (for purposes other than generating electricity). About 35 percent of all end-use energy (approximately 33,500 PJ) was electricity [7].

In 1995, the U.S. industrial sector used 36,376 PJ of energy--38 percent of total U.S. energy consumption (Figure 1b). Manufacturers consumed most of this energy (about 75 to 80 percent). (Nonmanufacturing industrial users in mining, construction, agriculture, fisheries, and forestry consumed the rest.) Electricity, including generating losses, accounted for 32 percent of the energy use (mainly to run motors). Natural gas accounted for approximately another 30 percent of energy used. In the manufacturing sector natural gas is used mainly for process heating, boiler fuel, and as a feedstock. In the industrial sector, including manufacturing establishments, petroleum accounted for 25 percent of the energy used. In manufacturing establishments, the petroleum is used mainly for process heating and boiler fuel [8].

Of all the energy sources used in the transportation sector, motor gasoline is used the most, accounting for 61 percent of the 25,397 PJ used in this sector. Distillate fuel accounted for 18 percent and jet fuel for 13 percent, of transportation energy use.

Electricity and natural gas are the main energy sources used in U.S. households. About one half of the electricity is used for appliances such as washers, television sets, etc. while most of the natural gas is used for space and water heating. In 1995, the residential sector used 20 percent (19,051 PJ) of the energy used in the U.S.

In 1995, the commercial sector used 14,711 PJ--the least of any of the four end-use sectors. The commercial sector uses mainly electricity, followed far behind by natural gas, (72 percent versus 22 percent). Almost one-half of the electricity is used for lighting while most of the natural gas is used for space heating [9].

          Energy Use in the United States: 1970 to 1995

    Before the 1970’s the United States experienced a time of falling energy prices and ample supplies of petroleum. As economic activity (measured by Gross Domestic Product (GDP)) increased, so did the consumption of energy (Figure 2). In 1973, crude petroleum prices shot up by 400 percent. At first short term effects, such as lowered thermostats and reduced driving, were common, but their overall effects on demand were small. While these transient actions were taking place, later to subside, other, more fundamental changes were working their way into energy-using processes. As these long-term changes in energy use characteristics and demand characteristics became permanent, the relationship between GDP and energy consumption weakened. The adoption of automobile and appliance standards, a shift in manufacturing away from energy-intensive processes, the growth of the service sector, and the introduction of energy-efficient appliances were just some of the more permanent changes [10]. In the early 1980’s the growth of economic activity outpaced the demand for energy.

    At the same time, a growing electrification, led by the commercial sector, has taken place (Figure 3) in the U.S. Electricity use grew by 53 percent between 1980 and 1995 in the commercial sector, and by 47 percent in the U.S. as a whole. With the growth of the service sector relative to manufacturing, the commercial building stock increased. In addition, the use of office equipment--from computers to copy machines--grew rapidly. In the residential sector the increased use of heat pumps for both space heating and air conditioning in the residential sector has been one of the factors leading to large increases of electricity use. This was especially true in the new homes constructed in the South, which experienced 30 percent growth in the number of housing units between 1980 and 1995. New homes also have been getting larger and households have been using more appliances, and multiples of the same appliances, such as televisions and computers [11]. Although the increase was not as large (13 percent), the industrial sector also experienced a growth in the use of electricity between 1980 and 1995. In this sector, there has been a shift from some of the more energy-intensive industries, such as the primary metal industries. This shift has not only been toward less energy-intensive industries, but also toward those that heavily use motors--the major use of electricity in the manufacturing sector [12].

    As the use of electricity has increased, there has been a strong push, with California leading the way, to restructure the electric utility industry. In March 1998 the Administration presented the Comprehensive Electricity Competition Plan. This plan proposes consumer choice of a power supplier by January 1, 2003. The speed of total restructuring or the effect of electricity restructuring on electricity demand is difficult to predict. Many debates are taking place on topics such as whether customers will see lower electricity prices, the fate of renewable energy, and whether energy-efficiency technologies will continue to be developed and implemented.

        Energy-Efficiency Indicators


There is no single commonly-accepted definition of energy efficiency. At one level, energy efficiency is a value-laden concept, referring to the relative thrift or extravagance with which energy is used to provide goods or services. From a more technical perspective, an increase in energy efficiency can be said to have occurred when either energy inputs are reduced for a given level of service, or there are increased or enhanced services for a given amount of energy inputs.

Energy intensity is defined as the ratio of energy consumption to some measure of demand for energy services. Energy intensity measures are often used to measure energy efficiency and its change over time. However, energy-intensity measures are at best a rough surrogate for energy efficiency. This is because energy intensity may mask structural and behavioral changes that do not represent "true" efficiency improvements such a shift away from energy-intensive industries. The choice of a measure of demand for energy services (a "demand indicator") in efficiency analysis is critical.

Although it is difficult to equate energy efficiency to a single intensity measure, or set of measures, some form of energy intensity is often the best we can do with available data. Indicators of energy intensity are useful, but we must remember that the underlying components are critical to interpretation. Without a structural context, the indicators can be misleading. The structural component of intensity is important because it shows where policy might or might not be directed.

When EIA asked participants in the energy-efficiency workshops to define "energy efficiency," participant definitions reflected two different perspectives: either a service perspective or a mechanistic, strict intensity, perspective. Some participants believed that energy-efficiency indicators could measure some kind of economic well-being, and suggested that a wide range of indicators would offer insight into the "ordinary business of life" and the relationships, causes, and opportunities in observed trends. One concept of efficiency is a strict technological (equipment-based) concept. This concept cannot be strictly measured by broad intensities, because intensities tend to carry structural and behavioral components. Alternatively, some participants believe that differentiating between intensity and efficiency is senseless.

The central question in the analysis of energy efficiency may really be "efficient with respect to what?" Measurement of energy efficiency always relates to the specific policy objectives at stake. Otherwise, why should we care how efficient we are? Are we concerned specifically about economic well being, higher productivity, increased employment and incomes, resource conservation, or improved environmental quality? Different answers call for different indicators. Consequently, the appropriate indicator is dependent on the policy objective. For example, if the policy objective concerned the environment, then the intensity indicator would involve carbon emissions. From the global warming perspective, the absolute carbon emissions are obviously most important, and energy intensity is not relevant. On the other hand, if economic productivity is the policy objective, then energy expenditures per dollar of GDP might be a more suitable indicator.

          United States Indicators

A variety of potential efficiency indicators exist for each end-use sector. (The next section (Section 5.3) will examine various indicators that have been proposed for the manufacturing sector.) However, whatever choices are made for individual sectors, at some point summary measures for the nation as a whole are needed. Obvious uses for national energy-efficiency indicators are policy formulation, assessment, and justification.

Measuring energy efficiency in the overall economy is even more complex than in the individual sectors. Measures of demand tailored to particular end-use sectors are likely to be inappropriate for the other sectors. EIA is considering two approaches toward the development of aggregate national indicators.

The first approach is to construct energy intensity measures using simple indicators of demand: population and GDP. Intensity measures constructed using these simple indicators can be useful for analysis. For example, the trend in energy use per person reflects historical energy price reactions, while energy use per dollar of GDP does not. The trend line for energy per dollar of GDP (Figure 4) shows a continuing reduction in energy consumption per dollar of GDP since 1970-- with a 19 percent reduction between 1980 and 1995. The story is different on a per-capita basis. Before 1980, and during the times of the "oil price shocks," short-term reactions to price changes can be clearly seen. Energy per capita after 1980 did not continue to decline but stayed rather flat between 1980 and 1995, reflecting favorable energy prices.

Besides the simple aggregated indicators demonstrated here, EIA is considering one further aggregation option, the construction of an energy-efficiency index. The relative nature of energy efficiency lends itself to the development of indexes. Individual sector indexes, suitably weighted, could be aggregated to form a national index.

Alternatively, a market-basket index, similar to the Consumer Price Index (CPI), is under consideration. A market-basket index starts with a fixed set of energy services. Compared over time to an engineering-improvement market basket of those same services, technical efficiency could be gauged. This approach captures the price-induced behavioral substitution shifts among basket items. However, a market-basket approach will share some of the same problems that the CPI does, such as the substitution effect--as the price of one good increases, the demand for a cheaper substitute increases also.

Manufacturing Sector Indicators

Two major users of industrial indicators are the academic and the research and development communities. Efficiency indicators may be used for policy analysis and some forecasting. An additional policy-oriented perspective could be considered, however, for identification of opportunities or target markets within the sector. Are we on a trend? Will opportunities present themselves or have we already found them all? Long-term industrial sector forecasting could use some type of end-use efficiency indicators.

Most indictors require a disaggregated approach, with more detail than 4-digit standard industrial classification (SIC). Ultimately, we need to look at indicators by process. However, the data needed to undertake such a highly disaggregated approach are unobtainable, either due to EIA's resource constraints or because the data are inherently difficult to obtain. Although the industrial sector includes mainly manufacturing industries, data on the other industries within the sector are virtually nonexistent. EIA does collect manufacturing data through the Manufacturing Energy Consumption Survey (MECS). The MECS data are collected every four years, but only by 4-digit SIC for the nine Census divisions. The only demand indicators with establishment data available for any underlying analysis are the indicators based upon MECS data. These indicators are the value of shipments, the value of shipments adjusted for inventory (value of production), and the value of production adjusted for capacity.

From an engineering perspective, efficiency measures should use physical measures of output, not economic value. However, although limited data are available for some industries, such as the aluminum or lumber industries, physical output data are not available for most industries. In the EIA database under development, we plan on including as many intensity measures using physical measures as possible. However, we will have to use a measure of economic value for most industries and especially for the U.S. as a whole. Economic values of output or demand will be problematic in an international context when trying to compare countries using different currencies.

In the development phase of this project we examined several measures of economic values as potential demand indicators: gross output (GO), gross product originating (GPO), industrial production, value added, value of shipments (VS), VS adjusted for inventory changes (value of production (VP)), and VP adjusted for capacity. Figure 5 shows examples of three of these, GO, GPO, and VS. Between 1988 and 1994, energy intensity did seem to fall, from 1 percent (using the GPO demand indicator) to 9 percent (using GO as the demand indicator).

In December 1996, a study comparing physical and economic measures was undertaken by the U.S. Department of Energy--Measuring Industrial Energy Efficiency: Physical Volume Versus Economic Value. Although this study advocates the use of physical measures wherever feasible, it states that the value of production demand indicator is the most desirable in the measure of energy efficiency. As the report states,

"Given that it is less likely to exaggerate swings in energy efficiency in the short run, and that it more closely matches growth rates in the long run than other value-based demand indicators, in the absence of serious coverage or specialization problems, value of production appears to be the most desirable value-based demand indicator for use in a measure of energy efficiency [13]."

EIA plans to use the value of production as the demand indicator. Additionally, the value of production will be adjusted for the changes in manufacturers’ internal mix of products produced during these years, and for the changes in the technologies and processes used to produce them. As a result of the workshop comments, the EIA energy-efficiency database will include indicators based on end-use data, primary and site energy, purchased energy and expenditures, and separate energy sources, by 2-digit and 4-digit SIC.

Transportation Sector Indicators

At the time that DOE was founded in 1977, U.S. energy issues were defined in terms of energy supply and energy security. Analysts were concerned that the U.S. had become too reliant on uncertain foreign sources of oil. Imports supplied 46.5 percent of petroleum consumption in 1977. Over a third of the petroleum consumed in the U.S. originated in OPEC countries [14].

After dropping to a post-1970's low of 27.3 percent in 1985, imports again supply almost half of the U.S. petroleum consumption [15]. However, the focus of energy-related concerns has shifted to the environment. Concerns about energy-resource depletion have largely been displaced by concerns about carbon emissions associated with energy use. The 1992 Rio Agreement and, especially, the 1997 Kyoto Protocol have heightened interest in the emissions of greenhouse gases.

To discuss emissions targets, a background understanding is needed on how energy is being consumed and the trends in energy consumption, as well as the associated emissions. The use of petroleum in transportation will serve as a simple example.

Since the 1970's, the role of petroleum has become narrower and more focused on serving the needs of the transportation sector. Electric utilities and, to a lesser extent, residential and commercial users, have moved away from petroleum as an energy source (Figure 6). Industrial use has remained fairly level, but represents a declining share of total petroleum consumption. However, the use of petroleum for transportation has more than compensated for declines in the other sectors. Transportation, which accounted for a bit over half of petroleum consumption in the 1970's, now accounts for over two-thirds. Ninety-seven percent of the energy consumed for transportation in 1995 was petroleum.

Overall, while transportation energy consumption per capita did not change greatly (up 6 percent) between 1980 and 1995, transportation energy consumption increased by more than 23 percent (Figure 7). The increase was not uniform. Transportation energy consumption peaked in 1978, at 21,145 PJ (95.2 GJ per capita). The 1978 level of consumption not reached again until 1986, although consumption per capita has so far remained below the 1978 level. Consumption has generally been rising (with a dip in 1991) through 1995 [16].

Carbon emissions have tracked steadily along with energy consumption. The trend lines for energy consumption and carbon emissions will remain virtually identical until nonpetroleum fuels emerge as significant sources of transportation energy. In 1995, the carbon emissions related to transportation energy use amounted to 457 million metric tons, almost a third of total U.S. emissions [17].

Several factors contributed to the increased demand for transportation energy (Figure 8). First, the U.S. population grew by 16 percent from 1980 to 1995. Second, the U.S. GDP increased by 46 percent over the 15-year period. GDP per capita rose 26 percent, so that Americans, on average, were wealthier. This increased wealth permitted an increase in purchases of sport-utility vehicles that consumed more energy per distanced traveled than do passenger cars. Finally, the number of motor vehicles in use increased by 27 percent. Motor vehicles, a large component of transportation energy use, are overwhelmingly petroleum-fueled. The number of motor vehicles per thousand persons increased from 713 to 778. Within the motor vehicle stock, the passenger car share declined from 80 percent to 67 percent, as light trucks became increasingly popular for personal transportation [18].

Population, GDP, and the number of vehicles all are gross indicators of the demand for transportation. More detailed measures include the distance traveled per vehicle and the quantity of energy required to travel a given distance. After dipping slightly in response to price shocks of the 1970's, starting from 1980, the distance traveled per year by the average motor vehicles increased steadily. However, vehicles used less energy to travel a given distance throughout the 1980's (Figure 9). After declining from the late 1970's to 1990, the trend toward less fuel used per unit distance flattened out in the 1990's. The two trends, toward more distance per vehicle and less energy per unit distance, have opposite effects on energy consumption and on the associated carbon emissions. Vehicles required less energy to travel a given distance, but they were driven greater distances.

This section has presented broad indicators of transportation energy use. These broad indicators do not do justice to the diversity of transportation sector (which includes both passengers and freight, and various modes of transport). Nevertheless, these indicators have led to some insight into the activities responsible for increases in transportation energy use and emissions. The indicators also highlight structural and behavioral trends in population growth, motor vehicle stock mix, and driving patterns. These trends need to be offset, either by efficiency gains or by changes in the transportation fuel mix, if carbon emissions in the transportation sector are to be reduced.


Recent developments have given new emphasis to the analysis of energy efficiency in the U.S. The Energy Policy Act of 1992 specifically requested the development of information on energy efficiency. Furthermore, on-going international negotiations regarding global climate change have highlighted the crucial role of energy efficiency in meeting possible treaty commitments without sacrificing lifestyles.

The Energy Information Administration (EIA), the independent statistical and analytical agency within the U.S. Department of Energy, began its energy-efficiency analysis in 1993. Initial efforts resulted in a methodological report (1995), a series of workshops (1996), and the establishment of an energy-efficiency area on EIA's web site (1996-present). EIA is now moving on to the next phase of energy-efficiency analysis, incorporating comments, suggestions, and experiences gleaned from the initial effort and subsequent research.

The U.S., with 5 percent of the world's population, uses 25 percent of the world's energy. Factors associated with this high rate of energy use include a high level of electrification, the world's highest GDP, and extensive land area. The U.S. consumed 95,300 PJ of energy in 1995. The industrial sector accounted for 38 percent of that consumption, followed by the transportation sector with 27 percent of the consumption. The residential and commercial sectors consumed the rest.

Several significant trends can be detected in U.S. energy consumption over the period from 1970 to 1995. Price shocks during the 1970's led to long-term changes in U.S. energy consumption patterns, and weakened the relationship between increases in energy consumption and increases in GDP. Except in the transportation sector, petroleum use has declined or remained flat, while electricity use has accounted for an increasing proportion of end-use consumption. It is too soon to tell the consequences of the restructuring of the electricity industry.

The definition and measurement of energy efficiency are not easy tasks. Energy-intensity measures are frequently used a surrogates. Energy-intensity measures have shortcomings, however, in that they can confound energy efficiency changes with other, structural and behavioral, changes. Ultimately, the choice of energy-efficiency indicators should be dictated by the purposes of the analysis.

The definition of indicators for individual sectors may be difficult, but further problems emerge when constructing indicators for the U.S. as a whole. Two approaches are being considered. One approach is to construct national indicators based on simple measures of demand, such population and GDP. A second approach would be some form of indexing.

Examples of the problems and uses of efficiency indicators are presented for two sectors, manufacturing and transportation. The manufacturing example illustrates the problems involved in choosing demand and efficiency indicators. Linkages among energy security issues, carbon emissions, and energy efficiency are demonstrated for the transportation sector.

In conclusion, EIA is taking a slow, deliberate, and thorough approach in the development of an Internet database of energy-use, energy-efficiency, and carbon-emissions indicators. By taking this approach, EIA will be able to produce indicators, within resource limits, that are not only robust, but also EIA will be able to update the indicators as new data are available.


1. "Energy Policy Act of 1992," Conference Report 102-1018, U.S. House of Representatives.

2. "Measuring Energy Efficiency in the United States’ Economy, A Beginning," pub. DOE/EIA-0555(95)/2, Energy Information Administration, October 1995.

3. "International Energy Annual 1996," pub. DOE/EIA-0219(96), Energy Information Administration, February 1998.

4. "Statistical Abstract of the United States 1995," pub. , U.S. Department of Commerce, September 1995.

5. "Annual Energy Review 1996," pub. DOE/EIA-0384(96), Energy Information Administration, July 1997.

6. "State Energy Data Report 1995," pub. DOE/EIA-0214(95), Energy Information Administration, December 1997.

7. "Annual Energy Review 1996," pub. DOE/EIA-0384(96), Energy Information Administration, July 1997.

8. "Manufacturing Consumption of Energy 1994," pub. DOE/EIA-0512(94), Energy Information Administration, December 1997.

9. "Lighting in Commercial Buildings," pub. DOE/EIA-0555(92)/1, Energy Information Administration, March 1992.

10. Heal, Geoffrey and Chichilnisky, Graciela, "Oil and the International Economy," pub., Clarendon Press, 1991.

11. "Housing Characteristics 1993," pub. DOE/EIA-0314(93), Energy Information Administration, June 1995.

12. "Changes in Energy Intensity in the Manufacturing Sector 1985-1991," pub. DOE/EIA-0551(85-91), October 1995.

13. "Measuring Industrial Energy Efficiency: Physical Volume Versus Economic Value", special report prepared for the U.S. Department of Energy by Pacific Northwest National Laboratory, December 1996.

14. "Annual Energy Review 1996," pub. DOE/EIA-0384(96), Energy Information Administration, July 1997.

15. "Annual Energy Review 1996," pub. DOE/EIA-0384(96), Energy Information Administration, July 1997.

16. "State Energy Data Report 1995," pub. DOE/EIA-0214(95), Energy Information Administration, December 1997.

17. "Emissions of Greenhouse Gases in the United States 1995," pub. DOE/EIA-573(95), Energy Information Administration, October 1996.

18. "Annual Energy Review 1996," pub. DOE/EIA-0384(96), Energy Information Administration, July 1997.

Back to the Conference PapersConference Papers blue_next.gif (2052 bytes) International Association for Energy Economics Paper



Specific questions on this topic may be directed to:
Stephanie Battles
(Phone: (202) 586-7237)
FAX: (202) 586-0018

Contact Us


File Last Modified: October 17, 1999