1. Introduction

Rankeda EI Numbers of Total Inputs of Energy
SIC Codeb	Intensity for 1985c	Intensity for 1994c
29	18.11	25.85
26	17.82	17.71
33	19.57	16.27
32	14.75	14.69
28	11.09	12.14
All	5.34	5.77
24	5.24	5.05
22	4.07	3.82
20	2.41	2.72
30	2.81	2.22
34	1.91	1.98
25	1.37	1.16
39	1.14	1.16
38	0.92	1.10
36	1.11	0.90
35	1.14	0.86
27	0.62	0.74
23	0.47	0.38
c For this report, all energy-intensity ratios are presented in units of thousands of Btu per 1992 constant dollars. Source: Table 12 of this report.

The focus of this data report is on energy consumption relative to constant dollar shipments of manufactured products -- commonly called energy intensities (EI) by energy analysts.  This report presents two measures of energy consumption, Offsite-Produced Energy and Total Inputs of Energy, while recognizing that structural and behavioral effects are enmeshed in those measures. Reporting EI represents the third edition of the Intensity Change report series.  While this report continues the previous reports, it does contain significant differences, such as the year basis for structural shifts and the treatment of electricity losses during generation and transmission. Specifically, this report:

• Presents adjusted estimates of energy intensity, removing the effect of different product mixes (structural shift) that were produced during each survey year.   With structural effects effectively removed, a clearer view of intensity changes is presented.  For the first time, structural shifts are calculated using more than one production year basis -- 1985, 1988, and 1991 product shipment mixes.
• Includes intensities based on 1992 constant dollars of value of shipments, assuming inventory changes between the beginning and ending years have an approximate net value of zero.  For information on a future release that will account for inventory changes, please click here.
• Presents energy-intensity ratios based on Total Inputs of Energy for heat, power, and electricity generation (including byproduct and waste products) as well as Offsite-Produced Energy.
• Excludes the use of primary electricity as a component in intensity ratios because of intractable data-quality issues in converting site electricity to include electricity losses during generation, distribution, and transmission.  For further information, please click here.

Energy-efficiency change cannot be inferred directly from changes in energy-intensity ratios. Improved energy efficiency, in and of itself, has the effect of reducing intensity.  However, other factors, such as shifts in production among industries and changes in use of onsite-produced energy, can also affect intensity, and in some cases will dwarf the effects of efficiency change.  This report adjusts intensity measures for both of these factors, so that changes in these adjusted measures might more closely represent the effects of energy efficiency.

Organization of This Report

This introductory chapter graphically summarizes changes of energy intensity and includes a discussion on the Standard Industrial Classification (SIC) system, which explains how the system identifies and classifies manufacturing establishments. Also discussed are the impacts of the 1987 SIC revision on data collected for each MECS, and how this revision is handled for longitudinal (i.e., cross-year) comparisons.

Discussions of two measures of energy consumption and manufacturing output are the focus of the second chapter. That chapter defines changes in energy intensity and illustrates the effects of using different definitions and comparing between different beginning and ending years. Also included in that chapter is a discussion of structural shifts in production and how these shifts affect measurement of energy intensity.

The third chapter contains detailed tables of SIC-specific information on energy intensities for each of the 20 major manufacturing groups (2-digit SIC) considered in this report. That chapter begins with a listing of 12 detailed data tables. Following these chapters are three appendices:

• Appendix A, Survey Design, Implementation, and Estimates, presents detailed information on the MECS design, implementation and estimates, and technical information on the calculation of estimates.
• Appendix B, Quality of the Data, discusses the quality of the survey data.
• Appendix C, Descriptions of Major Manufacturing Groups, includes descriptions of the major industrial groups covered by this report.

These appendices are followed by a glossary of statistical and engineering terms used in this report.

Manufacturing Energy Consumption Surveys

The Manufacturing Energy Consumption Survey (MECS) is the Congressionally mandated data collection that provides statistics for this report. Section 310(a) of the 1986 Omnibus Budget Reconciliation Act (Public Law 99-509, as amended) directs the Energy Information Administration to conduct the survey at least every three years.  Beginning with the 1994 survey, the frequency of the MECS has changed to every four years.  The next survey is scheduled to collect 1998 information during 1999.

The MECS focuses on consumption of all energy sources by U.S. manufacturers. It also collects establishment information such as fuel-switching capability and onsite fuel-storage capabilities. Additional enhancements to the latest released MECS may be found here.

Summary of Data

Data indicate a slight decline in the unadjusted amount of Total Inputs of Energy consumed per constant value of products shipped between 1991 and 1994. In 1985, manufactures consumed, on average, 5.34 thousand British thermal units (Btu) for every dollar of products shipped (1992 constant dollars).   This ratio inched to 5.44 in 1988, 5.51 in 1991 and settled at 5.49 in 1994. However, this trend of a slow rise in energy intensity comes on the heels of a precipitous drop in energy intensity.   Beginning in 1973, following the OPEC oil embargo, industry and Government joined forces to encourage the efficient use of energy. Such initiatives supported double-digit decreases in energy intensity.⁽¹⁾ However, those incentives created by industry and Government have decreased as more energy has become available and the price of energy has declined.  The extent to which price and availability affect energy intensity is a complex analytical issue and is best suited for discussion in an upcoming EIA article.

For a historical summary, tables 1-12 provide energy-intensity ratios and changes in  intensity for the every combination of the 1985, 1988, 1991, and 1994 surveys. For the changes between the latest two MECS (1991 and 1994), graphical summaries have been provided at the total manufacturing level and for each of the 17 2-digit SICs listed below.

Surveying the Manufacturing Sector

The manufacturing sector consists of establishments that use mechanical or chemical processes to transform material or substances into new products. An establishment is usually at a single physical location and is often called a plant, factory, or mill. It ordinarily uses power-driven machines and equipment for handling materials. Its products may be final products that consumers will purchase, such as an automobile or a chair, or they may be goods for use by other manufacturers, such as parts for automobile engines or rolls of upholstery fabric. A manufacturing establishment may also assemble parts or perform blending operations.

The Federal Office of Management and Budget (OMB) developed procedures for classifying manufacturing and non-manufacturing establishments into industry classes. These procedures are known as the Standard Industrial Classification (SIC) system. Each industry class consists of establishments that produce similar types of good or services.

Establishments that fall into SIC categories 20 (Food and Kindred Products) through 39 (Miscellaneous Manufacturing Industries) make up the manufacturing sector. Each of these two-digit major groups is further separated into three-digit industry groups and four-digit industries based on their primary products. For example, SIC 325 (Structural Clay Products) is one of nine industry groups in SIC 32 (Stone, Clay, Glass, and Concrete Products). SIC 3253 (Ceramic Wall and Floor Tile) and SIC 3255 (Clay Refractories) are two of the four-digit industries that make up SIC 325.

Revision of SIC Codes

OMB maintains and periodically updates the SIC system to ensure the classification system remains up-to-date, accurately reflecting current economic activities. Revisions consist of adding new categories as well as removing or combining old categories. Examples include creating separate categories for electronic computers and computer-storage devices, and combining categories for primary lead, primary zinc, and other primary nonferrous metals into a single category. The latest SIC revision was completed in 1997 with the introduction of the North American Industry Classification System (NAICS).   Fortunately, NAICS has not affected the estimates in this report because ALL intensity estimates have been placed on the 1987 SIC basis for comparability among MECS-sampled years.

SIC code revisions to industry groups and industries within a major group are reflected in changes in the third or fourth digit of a code. Moving industries from one major group to another causes changes in the first two digits. The 1987 version of the SIC codes was the basis for selecting establishments for both the 1988 and 1991 and 1994 surveys, while the 1985 MECS was selected under the previous version (1972). Estimates involving establishments that were reclassified at the level of the third or fourth digit but remained in the same two-digit major industry group are comparable with other estimates. However, when revisions moved establishments from one major group to another, the estimates are not comparable.  For consistency and analytical comparisons, energy consumption and value of shipments estimates are calculated for each SIC as though the 1987 SIC system were in effect for each MECS.

Changes between two-digit SIC categories must be considered when comparing data presented in this publication with data presented in earlier reports. For this report, all data conform to the 1987 SIC definitions. Because of these recalculations, direct comparison between this report and the reports associated with the 1985 and 1988 and 1991 surveys is neither advisable nor possible.

Industry Coverage

The MECS is a nationally representative sample of manufacturing establishments classified by SIC codes 20 through 39. In addition to these major manufacturing groups, SIC codes 01-07 (Agriculture), SIC codes 15-17 (Construction), and SIC codes 10-14 (Mining Activities) are outside the scope of the survey.  EIA plan to complete a comprehensive sector report on measuring energy efficiency. Please click here for further information.

The EIA constructed the MECS questionnaire and designed the sampling plan. The Census Bureau conducted the survey, assuring that each MECS response is covered by the same confidentiality provisions as apply to all Census Bureau data collections (Section 9, Title 13 of the U.S. Code). Response is mandatory. Only sworn Census agents have access to individual survey responses.

The sample design allows attachment of economic data obtained by other agencies (e.g., Census Bureau, Bureau of Economic Analysis). This matching is possible on an establishment-by-establishment basis. Data for this report include economic information for a specific establishment and energy consumption data for that same establishment. The Bureau of the Census and the Bureau of Economic Analysis (BEA) provide economic data. These data combine to provide the information required to develop energy-intensity estimates.

The MECS selects from the entire manufacturing sector through its relationship with the Census of Manufactures (CM) and Annual Survey of Manufactures (ASM) conducted by the Bureau of the Census. Appendix A provides further details on this relationship. The sizes of the MECS samples are as follows:

• 22,922 establishments for 1994
• 16,054 establishments for 1991
• 12,065 establishments for 1988
• 11,684 establishments for 1985.

This report, like previous reports, withholds estimates for some manufacturing groups. Three groups are omitted for confidentiality reasons and are identified as:

• SIC 21 - Tobacco Products
• SIC 31 - Leather Products
• SIC 37 - Transportation Equipment.
  

2. Energy Consumption and Production Output

Measures of Energy Consumption

This report presents information on two measures of energy consumption: (1) Total Inputs of Energy for Heat, Power, and Electricity Generation, and (2) Offsite-Produced Energy for Heat, Power, and Electricity Generation. Both measures provide specific information on how manufacturing establishments use energy.

Total Inputs of Energy for Heat, Power, and Electricity Generation is the more comprehensive measure of energy consumption. For the first time, this report includes intensities based on this measure of consumption. This measure:

• Includes all final uses of energy sources­produced either onsite or offsite­to produce heat and power, and to generate electricity.
• Includes net electricity, net steam and net industrial hot water (see glossary for further descriptions).
• Excludes energy sources used as raw material or for other nonfuel uses (e.g., coal used as an input to produce coke or wood used to make furniture).

Consumption of Offsite­Produced Energy for Heat, Power, and Electricity Generation is a more limited measure of energy consumption collected by the MECS. This measure:

• Includes all energy sources purchased or transferred from offsite and consumed onsite to produce heat and power and to generate electricity.
• Excludes all energy produced onsite and consumed as fuel.
• Excludes energy sources used as raw material or for other nonfuel purposes.

When Offsite-Produced is compared with Total Inputs of Energy measures, it shows how dependent particular industry groups are on energy purchased or transferred from offsite sources. For example, the Paper and Allied Products Industry (SIC 26) consumed 19.36 thousand Btu for every 1992 constant dollar value of output in 1994. Of this Total Input, only 11.91 thousand Btu came from Offsite-Produced Energy. Much of the remainder of the energy consumed was in the form of pulping (black) liquor, a combustible byproduct of the chemical pulping process.

This report excludes information on electricity losses during generation and transmission to manufacturing establishments.  The previous intensity-change report included information on energy intensity when the primary conversion factor for electricity is used rather than the site conversion factor. This measure of consumption includes the amounts of energy required by electricity generators to produce electricity consumed by manufacturers, and the amount of electricity lost in transmission and not available to the manufacturing establishment.

Value of Shipments as Output Indicator

The calculation of changes in intensity requires a measure of production output for each SIC. One possible measure is the number of physical units produced; however, physical quantities are not the best measure for two reasons. First, most manufacturing establishments produce more than a single product line. For example, in SIC 22 (Textiles) a single establishment may produce not only bedspreads, but also tablecloths and dishcloths. Because these are different products, with different process steps and energy requirements, summing these physical units to a single measure of total output is not possible. For the same reason, it is not possible to sum the physical outputs of all industries to obtain an estimate of total manufacturing sector output.

Second, summing different products into a single output indicator might be appropriate if information on energy consumption associated with each product line were available, but manufacturers rarely, if ever, monitor and record energy data by product or process. This information could suggest the use of a particular production process, possibly affecting their competitive status. An associated difficulty is that the MECS and ASM obtain data by establishments, not by products. Combining information from different sources would be nearly impossible if product rather than a controllable economic variable represented output.

Value of shipments is the output measure used to calculate energy intensity. Summing value of shipments both across and between SIC categories provides output measures for each SIC and for the manufacturing sector as a whole. Value of shipments is also the production measure used by the Census Bureau when assigning establishments into a particular SIC category. Use of this measure further assures consistency between the MECS and the ASM.

There exist other economic measures for intensity analysis. A potential competing output measure is value added (see glossary). If the goal were to calculate the total output of the entire manufacturing sector, value added would be the appropriate economic measure to use because it does not duplicate economic inputs as does value of shipments. The duplication inherent in the value of shipments is not an important consideration when calculating changes in intensity. The purpose is not to compare the energy-intensity ratios between industries, but rather, to compare ratios at different points in time for each manufacturing group. For further details see Appendix A.

Using value of shipments as an output indicator also has drawbacks. One problem is that the product values change over time. Inflation can cause these changes. Rising values of shipments over time are not necessarily associated with increased production but instead could reflect general price increases that affect all products and services to varying degrees.

The value of shipments estimate used in the calculation of energy-intensity ratios is adjusted for fluctuations in product value over time. Using BEA price deflators at the industry level, each individual establishment's value of shipments for 1985, 1988, 1991, and 1994 is adjusted to represent 1992 constant dollars.⁽²⁾ By adjusting for changes in price of goods, aggregate output measures of value of shipments are made comparable over the time periods of interest.

Further, re-estimation of value of shipments statistics is required to ensure comparability with estimates of energy consumption. Rather than adjust the value of shipment estimates provided by the ASM, this report uses the direct estimates of value of shipments from the MECS sample, which required a 100 percent matching of each individual MECS establishments with their corresponding ASM value of shipments. Direct estimation of both consumption and value of shipments statistics from each of the MECS samples results in comparable energy-intensity ratios.

A second problem with using value of shipments as an output measure is that, over time, the contribution of each SIC to total output may vary. This variation can occur within and between two-digit groups. What might at first glance appear to be an increase or decrease in energy intensity could, in fact, be simply a response to changing demand for particular products. Structural shift is the term that describes these changes in relative importance. (See box for a hypothetical example of structural shift.)

To estimate structural shifts, EIA developed another set of intensities for each SIC for both Offsite-Produced and Total Inputs of Energy consumption, which are presented in Tables 3, 4, 7, 8, 11, and 12. The structurally adjusted intensity change is based on the assumption that no structural shift occurred since the base year (i.e., the earliest year in the comparison), and that the same mix of output was available in the ending year.  For example, comparisons between 1985 and 1994 use the base year 1985, while comparisons between 1991 and 1994 use the base year 1991.

Changes in Energy Intensity

Manufacturers are in the business of producing physical units of output for consumption by end users or other manufacturers. One goal of production is to consume as few inputs as possible to produce a quality output. Specifically, if energy requirements are reduced relative to output, each unit of output has become less energy intensive. This improvement is an increase in energy efficiency.

This report presents the percent changes in energy intensity for every combination of all MECSs taken two years at a time: 1985-1988, 1985-1991, 1985-1994; 1988-1991, 1988-1994; and 1991-1994. This information is presented for all 20 two-digit major groups, except Tobacco (21) and Leather (31). For Tobacco and Leather, sample sizes for earlier surveys were not sufficient to produce representative estimates. This report includes 1985 and 1988 survey data recalculated to include all establishments in their 1987 SIC classification. Because of the change in the valuation of product shipments as 1992 constant dollars, comparisons between published reports are not possible.

Like previous reports, this report calculates changes in intensity as the difference between the energy requirements per constant dollar value of shipments for one period and those for an earlier period. The energy-intensity ratios included in Tables 1 through 12 provide the information to calculate the changes in intensity. These ratios indicate Btu consumption (in thousands) for each constant dollar of output. A percent change in efficiency that is positive indicates an increase in energy efficiency, whereas a negative percent change indicates a decrease in energy efficiency.  Similarly, an increase in the ratios of energy intensity from one year to a subsequent year denotes a decrease in efficiency, while a decrease in the ratios indicates an increase in energy efficiency.

It is important not to compare changes in intensity between different SICs but instead to compare the changes within an individual SIC. The output from one manufacturing establishment is often used as an input by another establishment. To some extent, the value of shipments of the second seller duplicates that of the first seller. For this reason, the sum of individual SIC value of shipments will not result in the value of shipments for the entire manufacturing sector.

Structural Shifts

Energy intensity is a product of both the efficiency of the manufacturing process and the mix of products being produced. Over time, new products (e.g., personal computers, compact discs) become available and replace older technology (e.g., typewriters, phonograph records). Analysis of structural shift identifies this change in demand, and the impact it has on manufacturing establishments. Energy consumption by an establishment or major group changes in response to increased (or decreased) demand for products. This report identifies the importance these changes in relative output have on energy intensity. Tables 3, 4, 7, 8, 11, and 12 identify Offsite-Produced and Total Inputs of Energy intensities adjusted for structural shift.   For the first time, this report applies more than one basis for structural shifts -- 1985, 1988, and 1991 production mixes.⁽³⁾

Appendix A provides a mathematical description of how structurally adjusted energy-intensity ratios are derived. As the 1988 or 1991 or 1994 unadjusted intensities are applied to a given year's output distribution, the result is the amount of energy that would be required to produce the same distribution of products, but at the later year's energy intensities. If an adjusted intensity is less than the unadjusted intensity for that same year, this indicates a movement within the two-digit group toward more energy-intensive products over time. At the same time, if the adjusted intensity is greater than the unadjusted intensity, movement within the group was toward less energy intensive products. Comparison of unadjusted and adjusted percent changes in energy intensity indicates how much of the unadjusted change was due to changes in efficiency and how much was due to structural shifts within manufacturing.

Chart Interpretation

Energy Intensity Chart

The summary charts presented for 18 of the 20 2-digit SIC present both the unadjusted and structurally adjusted energy-intensity ratios, along with an annualized percent change in efficiency, as measured by changes in adjusted energy-intensity ratios. Horizontal bars represent those energy-intensity ratios for both Offsite-Produced Energy and Total Inputs of Energy. The actual ratio for the given year is adjacent to the bar. All estimates for consumption and value of shipments were derived directly from the MECS samples.

Given a base year and end year, the annualized percentage changes in efficiency for both Offsite-Produced and Total Inputs of Energy are displayed adjacent to their respective energy-intensity ratios. A decrease in energy-intensity ratios denotes an improvement in energy efficiency, while an increase denotes a loss in efficiency.

Both structurally adjusted and unadjusted energy-intensity ratios are shown. Because 1991 serves as the base year for structural-shift adjustments, energy-intensity ratios adjusted for structural shifts are shown only for the 1994 end year. If the structurally adjusted ratio is less than the unadjusted ratio, a movement toward more energy-intensive products since 1991 is indicated. Similarly, when the structurally adjusted ratio is more than the unadjusted ratio, a movement since 1991 toward less energy-intensive products is indicated. An absence of movement (or a net movement of zero) is indicated by equal adjusted and unadjusted ratios.

Energy-intensity ratios adjusted for structural shifts can provide relatively clean measures of efficiency change. However, such measures, if they are based only on Offsite-Produced Energy consumption, are incomplete because they ignore the potential impacts of the consumption of Onsite-Produced Energy, such as black liquor, wood chips, petroleum coke, and waste gas. Of the 16.52 quadrillion Btu of Total Inputs of Energy consumed in 1994, byproduct consumption accounts for 4.51 quadrillion Btu (27 percent). If Onsite-Produced Energy sources serve as substitutes for Offsite-Produced Energy sources, a change in the relative amounts of the two sources used could mask changes in energy intensity.

3. Detailed Tables

There are 12 data tables used as references for this report.  Specifically, these tables are categorized as

• tables 1 and 2 present unadjusted energy-intensity ratios for Offsite-Produced Energy and Total Inputs of Energy for 1985, 1988, 1991, and 1994; along with the percentage changes between 1985 and the three subsequent years (1988, 1991, and 1994)

• tables 3 and 4 present 1988, 1991, and 1994 energy-intensity ratios that have been adjusted to the mix of products shipped from manufacturing establishments in 1985

• tables 5 and 6 present unadjusted energy-intensity ratios for Offsite-Produced Energy and Total Inputs of Energy for 1988, 1991, and 1994; along with the percentage changes between 1988 and the two subsequent years (1991 and 1994)

• tables 7 and 8 present 1991 and 1994 energy-intensity ratios that have been adjusted to the mix of products shipped from manufacturing establishments in 1988

• tables 9 and 10 present unadjusted energy-intensity ratios for Offsite-Produced Energy and Total Inputs of Energy for 1991 and 1994; along with the percentage changes between 1991 and the subsequent year (1994)

• tables 11 and 12 present 1994 energy-intensity ratios that have been adjusted to the mix of products shipped from manufacturing establishments in 1991.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

W = Withheld to avoid disclosing data for individual establishments.

Note:  Positive percent change indicates a decrease in energy intensity and an increase in energy efficiency.  Negative percent change indicates an increase in energy intensity and a decrease in energy efficiency.

Source:  Energy Information Administration, Manufacturing Consumption of Energy 1994, DOE/EIA-0512(94), Washington, DC, 1997; Manufacturing Consumption of Energy 1991, DOE/EIA-0512(91), Washington, DC, 1994; Manufacturing Energy Consumption Survey:  Consumption of Energy, 1988, DOE/EIA-0512(88), Washington, DC, 1991; and Manufacturing Energy Consumption Survey:  Consumption of Energy, 1985, DOE/EIA-0512(85), Washington, DC, 1988.

Endnotes

The energy intensities presented here and in past energy intensity reports are based on only value of shipments as the output (or consumer demand) indicator and do not reflect changes in beginning- and ending-year inventories. Energy intensities adjusted for changes in inventory will be available at a later date as well as an analytical review of these energy intensities.  For more information on EIA's efforts to measure energy efficiency, please click here.  For international comparisons, please access The Handbook on International Comparisons of Energy Efficiency in the Manufacturing Industry, Phylipsen G.J.M., Blok K., and Worrell E., Department of Science, Technology and Society; Utrecht University.

The correct terminology for the chain-weighted, industry-specific deflators used in this report is Chained (1992) Dollars.  For consistency with a previous EIA report - Manufacturing Consumption of Energy 1994 (DOE/EIA-0512(94)) - the term "1992 Constant Dollars" is used in the place of "Chained (1992) Dollars".

1. Energy Information Administration, Changes in Energy Intensity in the Manufacturing Sector 1985-1991, Figure ES2, page vi, DOE/EIA-0552(85-91).

2. U.S. Department of Commerce, Bureau of Economic Analysis, National Income and Wealth Division.

3. All MECS surveys are subsampled from the sample for the Annual Survey of Manufacturing (ASM) conducted by the Bureau of the Census. The ASM routinely produces estimates of value of shipments by three-digit manufacturing industry groups. Because of the subsampling procedure used for the MECS, it is possible to replicate those value of shipments estimates by using MECS establishment weights applied to ASM-reported value of shipments for respondents that were in both surveys. If the MECS-weighted estimates of value of shipments were within 10 percent of the ASM estimate for any given three-digit industry group, that group was judged to be sufficiently reliable to assess structural shift. Those estimates with a difference greater than 10 percent were combined until the 10-percent criterion was met. Counting single three-digit industry groups, 83 groups were judged to be sufficiently reliable when the comparison years include the 1985 or 1988 MECS and 98 groups are reliable for comparisons based on the 1991 MECS.

a. Rankings of intensities are based on the adjusted 1994 energy-intensity ratio of Total Inputs of Energy.

b. SICs 21, 31 and 37 are excluded from this table because data were withheld from the 1994 MECS because of conflicts with data confidentiality (Title 13 of the US Code).  Note that both output and energy consumption estimates for SICs 21 and 31 are contained in higher aggregate totals.

Top of Page
Energy Intensity Page
MECS Home Page