- Email: [email protected]
Uses of economic census data
Uses of Economic Census Data
GAYLORD WORDEN
The comprehensive economic censuses conducted by the Bureau of the Census at 5-year intervals are the foundation of the nation’s economic statistics programs. These censuses play a critical role in allowing millions of private and public decision makers to make more informed economic plans and decisions. The uses of these data are limited only by the failures of human imagination. This article describes briefly some of the principal uses of the economic censuses. The United States economy is large and complex. Millions of private and public sector
decision makers are involved daily in keeping it healthy and vigorous. They rely on information from the nation’s economic statistics programs to better understand the complexities and reduce the uncertainties surrounding the decisions they make. The economic censuses, conducted once each 5 years, are the foundation of the nation’s economic statistics programs. They are the most comprehensive collection of data on the many sectors of the economy. Because these data are consistent, reliable, and detailed, they are used in a myriad of ways to address the structure and performance of the economy. They are used to address detailed questions on inputs used and outputs produced by industry or kind of business, both nationally and in small geographic areas. And they provide controls and sampling frames for surveys that produce more current but less detailed measures of economic activity. This article describes the major uses of data from the economic censuses. These uses are: l l l l l
framework for current economic surveys; measuring and tracking changes in economic activity; calculation of composite measures of the nation’s economic activity; analysis of economic policies; planning and monitoring of economic programs; and planning and management by private sector businesses.
Gaylord Worden is Chief; Industry Division, U.S. Bureau of the Census, Washington, D.C. 20233. Government Infommtion
Quarterly, Volume 4, Number 3 (1987), pages 245-259.
ISSN: 0740-624X.
246
GOVERNMENT
INFORMATION
Before these uses are discussed in some detail, a description content of the economic censuses is provided.
A DETAILED
SNAPSHOT
QUARTERLY
Vol. ~/NO. 311987
of the basic structure
and
OF THE ECONOMY
The economic censuses often are described as providing a detailed snapshot of the economy. This snapshot is taken for years ending in “2” and “7.” It is the comprehensive coverage of economic sectors, the industry or kind of business and geographic area detail, and the reliability and availability of the data that make the censuses so useful. No other statistical program provides similar data on economic activity on such a scale. Most sectors of the economy are included in the economic censuses. There are separate but coordinated censuses for manufacturing; mining; wholesale trade; retail trade; business, personal and professional services; and construction, Some transportation sector industries are covered while public utilities, finance, insurance, and real estate are the principal sectors not included. Regulatory programs traditionally have provided data for these sectors. Agriculture and governmental units are covered in separate but parallel censuses.
EstablishmentData For each of the economic sectors covered by the censuses, each plant, store, or other business location with employees is identified and included in the censuses. Each location is commonly called an establishment. The establishment may differ from the company or firm, which is an organization consisting of one or more establishments under common ownership or control. Nonemployer establishments are also included in the retail trade, services, and construction censuses. For each establishment, data are collected on employment, payroll, and other inputs and costs. Comparable data on shipments, receipts, and other measures of output or production help complete the snapshot of economic activity. The amount of detail on inputs and production varies significantly by sector. The most detail is provided for manufacturing where data on nearly 1,000 input materials and over 11,000 products produced are collected in the census. Data for other sectors generally are much less detailed. Based on data reported on the primary products produced at a manufacturing plant or the major kind of business conducted at a business location, the establishment is classified using the Standard Industrial Classification (SIC) system into the 4-digit industry.’ This system is used widely for statistical data by Federal agencies as well as the private sector to provide consistency in the way data are classified and to allow users to consult data from many sources. Data from the economic censuses provide detailed information by industry because they are classified at the 4-digit level. Each establishment also is assigned a detailed geographic code. This code allows it to be classified by city or place, county, metropolitan area, and state. Again, no other statistical programs provide such comprehensive and consistent economic data for these small geographic areas for the entire nation. Table 1 (pp. 248-249) shows the major data items collected in the censuses by geographic level of detail.
Use of Economic
Census
Data
247
Because data are collected by establishment and classified by a detailed industry description and geographic location, establishment data are often called building blocks for statistical information. These building blocks provide: comprehensive coverage of most of the economy; consistent basic input and output, or production data across the economy; l data for detailed industry classifications; l data for small geographic areas; l great flexibility in how the data can be aggregated and used; and * another data point each 5 years that is consistent with an already long series of data.
l
l
EnterpriseData An additional building block included in the censuses is enterprise-(company) level data. By assigning a common identifier to all establishments of the same company, the Census Bureau is able to put together all the establishments for companies, providing a rich set of data for analyzing structure and organization of economic activity. Making this linkage between information for enterprises and establishments available once each 5 years also allows data users to make better use of the variety and combinations of independent establishment and company-level data available in other years. For example, data used in calculating the Gross National Product (GNP) accounts quarterly and annually include establishment data from many surveys as well as statistical information from company-level reports filed with the Internal Revenue Service. Other Data In addition to the building blocks described above, other programs are included in the economic censuses to provide selected data once every 5 years. One sample survey collects inventory and use information on the nation’s fleet of trucks. A special program provides data on minority-owned business enterprises, and another provides data on women-owned businesses. Still others provide detailed data on water use by manufacturers and mining operations, and on the types of textile machinery being used. PRIMARY USES OF CENSUS DATA Only the failures of human imagination limit the uses made of data from the economic censuses. Because these data measure the structure and activity of a large portion of the economy on a consistent and scheduled basis, they constitute a major body of facts and information that are used at all levels of government and in the private sector. The primary and most frequent uses are described in this section. But because the ways the economic census data might be used are truly limitless, they are so widely available,2 and there are so many users, it is not possible to chronicle all users and uses.
.. .. .. .. .. .. . . . . 0
*
248
.
.
.
.
.
.
.
*
*
.
.
1
.
.
.
.
‘
*
*
*
.
*
*
*
.
.
.
.
.
*
.
.
.
.
.
.
.
*
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
=
.
.
-
.
.
.
.
.
... ... .. .. . . .*.
.. .. ..
.
.
.
*. .. .. 1 .
I
.
.
.
.
.
.
.
.
.
.
.
.
*
.
.
f
*
.
I
.
*
.
* *
._
.
.+a
.
. . f .
. , .
. * .
.
.*4J
. . it! . .k.6h D . “* . * * :zJa \it -1. 3 0. .l E :. .9-r :,oli 5 u: :trP p : :?:: Ia. * :2; ?I : :2,” $ : .*ai: I
.h .*, ..4
.e .C . .
r,
. .m .
.” . . .
.
.
.
.
. ..I-
:
SO .\
. .
.m . .
. :Jr; !z w. ‘“J
.a .o
.
.. ... .. . ... ..
... ... ... ... .
.. .. .. .. .. ..
.
.
.. .. .. .. ... .
.* ... .. .* ...
1
.
.
*
*
.
.
*
*
.
*
*
*
.
.
.
I
.
*
.
*
.
.
.
.
.
.
.
.
*
*
.
.
.
.
.
.
.
.
.
.
::
ii
.
* . . .
::
i5
:L :: .h . :B :%S . . :4 ;g ;g :;’ :=i
;
:;+” ** 0 *alo
*c
::
G’XSS
$2
249
.. .. .. i.. ...
2
250
GOVERNMENT
Framework
INFORMATION
QUARTERLY
Vol. ~/NO. 3/l 987
for Current Economic Surveys
Since the censuses are only conducted once each 5 years, it is necessary to have more frequent sample surveys to measure economic activity and to update the 5-year censuses. There are annual sample surveys of manufacturing, retail trade, wholesale trade and services. Important monthly surveys3 also are conducted for manufacturing and retail trade. The censuses provide the best available and most complete list of business establishments, and key information about their activities. The list is needed to draw the sample panels for a survey. Being able to draw a new sample once each 5 years from an up-to-date list helps assure the accuracy of these sample surveys. With the rapid changes taking place in the economy, less frequent development of new samples would lead to problems in the accuracy of these current measures of the economy. Some of the current sample surveys also are benchmarked to or reconciled with the censuses.4 Since the complete coverage in the censuses provides more accurate data than the surveys that are subject to sampling error, the benchmarking process improves the sample estimates. In a similar way, some trade associations benchmark data they collect to the censuses. Measuring and Tracking Changes in Economic Activity
The comprehensive and detailed data from the economic censuses are used by all types of users to address a myriad of questions about changes in economic activity. Because these building blocks of data can be put together in so many ways, they are highly useful for such analytical activities as charting long-term trends in the economy; identifying what industries and businesses exist, their relative importance, and their geographic location; and measuring what kinds of material they consume, how much labor they employ, and how much capital they require. A few examples of broad questions that can be answered are: l
l
l
l l l
l
l
What is the changing mix of manufacturing industries and which are growing or declining most rapidly? Are multiestablishment companies becoming more vertically integrated or horizontally diversified? Are department stores accounting for a smaller or larger share of total clothing sales? What has been the shift in textile employment by geographic regions (see Figure l)? Are small firms increasing their share of road building activity? Which states have the biggest increases in business services activity? Has the ranking of states by level of business services receipts changed in the past 10 years? Do minority-owned businesses account for a larger share of economic activity than they did 10 years ago? In which types of economic activity are minority-owned businesses increasing their shares most rapidly? In which states are Hispanicowned businesses increasing most rapidly? Have manufacturing sales branches and oftices increased their share of wholesale activities since 1972?
Figure 1.
100
Percent Textile Employment
in New England and South Atlantic
States.
Percent a m
New England South Atlantic
90
80
60
En-
30 24
20
10
1963 Year 251
1972
1962
Figure 2.
Share of Value of Shipments
Four and Eight Largest Companies Percent (Concentration
in the Electronic
Accounted Computing
for by the Equipment
Industry.
ratio)
100 m m
Four Largest Companies Eight Largest Companies
90 83 80
1967
1972
1977 Year 252
1982
Use of Economic
Census Data
253
Other questions may focus on a particular geographic area, kind of business, or product produced, or the relationship between two industries or two economic variables. A few examples are: l
l l
l
l
l
l
l
l
What changes are occurring in manufacturing employment in the Cleveland metropolitan area? Is employment growing in some primary metals industries and declining in others? Are food store receipts higher in Waterloo or Cedar Rapids, Iowa? Were receipts the same in 1977? How does the change in manufacturing activity in North Carolina compare with other southeastern states? What are some of the products in which aluminum is being used more? What materials is it replacing? Is the manufacturing of electronic computing equipment becoming more concentrated (see Figure 2)? Is retail activity in the center of Ft. Worth growing or declining relative to surrounding suburban areas? How has the mix of labor and material costs changed in the past 15 years for the manufacturers of dairy products? How specialized is the wood furniture industry?
These are a few examples of what can be an endless stream of very diverse questions. All potential users of the data, whether in a local, state, or Federal government organization or in a small or large business, ask these types of questions about economic activity. Only the economic censuses provide the broad but detailed data required to answer such a wide variety of questions. Some of the broader questions listed above also can be answered with data from annual or more frequent sample surveys. But such an ability to measure and track trends in economic activity with these more current surveys is dependent on the economic censuses, which provide the framework for the current surveys. Composite Measures of the Nation’s Economic Activity
Some of the best known economic statistics series are those that provide current composite measures of economic status and trends. These include quarterly estimates of GNP and nonfarm productivity, and monthly series such as the Index of Industrial Production and the Producer Price Index. The economic censuses are critical in maintaining accuracy of each of these. Benchmarking the national income and product accounts (NIPA), including the level of GNP and all its components, to data from the economic censuses is an extremely important process. Since no other data provide the detail and accuracy of the economic censuses, the GNP estimates, benchmarked to the censuses, are the most complete and accurate possible. This benchmarking is a very detailed and involved process carried out by the Bureau of Economic Analysis (BEA).
254
GOVERNMENT
INFORMATION
QUAKTERLY
Vol. ~/NO. 311987
The first step in this process is to use data from the censuses as well as many other sources to prepare a detailed picture of the economy showing the production and use of commodities and services among industries. This picture, known as input-output tables, is prepared by tracing how each of the economy’s industries use the products of all industries in producing final products and services for consumption or investment by other industries. This requires the detailed product and material data collected in the censuses as well as information on sales by class of customer which is collected in every other census or once each 10 years. These tables are useful for analyses of how changes in demand for consumer and investment goods, government expenditures, exports, and imports affect all industries. These updated input-output tables then are used to derive benchmark estimates of the NIPA. All of this is dependent on the data from the economic censuses. While the input-output tables use data from all the economic censuses, the manufacturing and mining industries comprise about 70% of the commodity classes covered by the tables. These complex tables are prepared only for economic census years when the detailed data are available. The Bureau of Labor Statistics (BLS) and the Bureau of Economic Analysis (BEA) also use data from the censuses to update the weights assigned to various products in important price series. The BLS uses data from the Census of Manufactures on the quantity of products shipped to determine how much weight to assign to the price of each of several thousand products included in the Producer Price Index in order to summarize them into more aggregate index numbers. In a similar manner, BEA relies on census data for its price series maintained for deflating GNP estimates. Updated weights from the censuses5 help maintain the accuracy of price series by reflecting the current mix of products being produced. Accuracy of price series is, in turn, vital for accurate estimates of real economic activity, that is, estimates adjusted to reflect changes in prices. A third major way census data are used as the foundation for current composite measures of economic activity is for benchmarking and weighting the monthly Index of Industrial Production by the Federal Reserve Board and the quarterly estimates of nonfarm productivity by BLS. Periodic updating of weights for these series are required because of the changing mix of industries over a 5-year period. And as this mix changes, the Ievels of these series can be corrected only by benchmarking to the economic censuses. Providing this foundation for the composite economic measures is one of the more important uses of the economic censuses data. Many decisions on the content and scope of coverage of the censuses are made based on how that decision will affect their use in composite measures. Having accurate composite measures is critical for the development of informed public economic policies and for private decision making. All investors, consumers, wage earners, and producers are affected by the accuracy of current composite measures and their contribution to reducing the uncertainties surrounding the decisions they have to make. Analysis of Economic Policies: Planning and Monitoring
of Economic Programs
Many economic policies and programs require more detailed information than what is available from composite measures of the national economy. Frequently the greater
2.55
Use of Economic Census Data
detail required is by small geographic area; sometimes it is by industry or product. The economic censuses serve these needs very well. A common use is for planning or monitoring economic development programs for a local area. Or the data may be used to assess the geographic distribution of economic activity or changes in that dist~bution to determine the need for regional economic assistance programs. Another important use is to develop plans and programs for the nation in the event of an emergency. Others want to evaluate the performance of programs to encourage the development of minority-owned businesses or small businesses. Another example is the use of concentration ratio data to judge the performance of existing antitrust programs or the need for additional programs. Only data available from the economic censuses once each 5 years can serve all these purposes. These are examples of policies and programs that are common to all levels of government. State and local bodies make frequent use of the data by small geographic area. The censuses give them a rich, though infrequent, source of data for decisions on how to provide efficient services and plan for economic growth. These same users in all levels of government rely on the less detailed data from annual and more frequent surveys for analysis of economic policies and programs. Since the accuracy of these more current data is vital for these uses, economic censuses also are critical in this indirect way. Planning and Management
by Private Sector Businesses
While the uses of census data already described are important to private sector business managers, we need also to describe and highlight some of the specific and direct ways the private sector can use census data. It is important to recognize that many of the suppliers of the data, the manufacturer, the retailer, or the service establishment also are among the primary users of the results. The more knowledgeable they are concerning how census data can benefit them, the more likely they are to participate in the censuses and surveys and to respond accurately and quickly. A major task facing any statistical agency is to educate its suppliers concerning how they can use the data. Since not all firms are aware of the direct benefits to be realized from census results, some examples of how they can use the data to improve both production and marketing decisions and strategy are useful. /~~roy~ng
P~ucfjon
Strategy
Census data can be used by individua1 firms in planning and developing their production strategy. In many cases, this strategy depends on assessing the performance of their firm against other firms in the same industry or geographic area. By comparing itself with those in the same general category, that is, in the same industry, same geographic area, same relative size class, or those specializing in the same types of products, the firm compares its own performance to the industry as a whole. For example, using the latest data from the Census of Manufactures or the Annual Survey of Manufactures series, the firm can compare its performance to certain industry averages. Averages or ratios that may be used for comparison include the following: l
payroll per dollar of sales;
256
GOVERNMENT
INFORMATION
QUARTERLY
Vol. ~/NO. 3/1987
l cost of materials per dollar of shipments; 0 payroll per employee; l annual wages per production worker; l production worker wages per hour; l nonproduction workers to production workers; l capital expenditure per employee or dollar sales; and l unit value (value/quantity) for individual products.
The various measures that could be considered for judging the performance of the individual operating unit can be extended to as many data items as are collected. As part of the Census of Manufactures, the reports provide many ratios for each industry over a number of years in order to make firm/industry comparisons easy. By comparing changes from year to year for products, industries, or geographic areas, with changes that have taken place in the firm itself, the company can determine whether it is improving its performance relative to the rest of the industry. If its performance is unduly different, the firm can investigate the possible causes. Using the census data, the firm can compare itself to other firms classified in its industry in terms of the mix of specific materials consumed or the mix of labor and materials’ costs. If its mix is noticeably different, and the firm has been losing its market share, it may want to investigate why. Other comparisons useful in production planning can be made from census data. For example, a table from the Census of Manufactures breaks down the value of industry shipments according to the primary and secondary products that are typically produced in the industry. Some of this industry product mix evolves naturally as a by-product of the manufacturing process. Secondary production may be a logical extension of using some of the waste from an earlier stage of production. In other cases, secondary products have evolved because individual firms have found it advantageous to produce commodities that can be marketed together with their primary product. The individual firm can examine how its product mix compares with other firms in the industry and examine whether there may be a more effective product mix. Improving
Market Strategy
Census data also permit the individual firm to evaluate its position in a market or determine the existence of a potential market. For example, the individual firm could analyze its share of the production of an individual product in the United States or in a specific state. Or the firm may look at its share of a broad class of products produced in a particular state. If this information is too aggregated, individual firms frequently request that the Census Bureau provide special tabulations of individual products produced or individual materials consumed by manufacturing establishments according to geographic areas that fit their marketing or sales district. If the company produces a product that is used in other manufacturing processes, it can determine the industries that are most likely to use its product, and by studying the geographic location of those industries, identify the most fruitful marketing areas. If the firm is producing a product that is used by or distributed through other economic sectors, such as retailing, wholesaling, or construction, it can use data from
Use of Economic
Census Data
257
the economic censuses for those sectors to identify the best potential markets among these types of businesses. Similarly, a firm can use this information to lay out sales territories and to pinpoint markets for advertising purposes. Information on market potential and potential suppliers, if used in conjunction with knowledge of the demographic characteristics of a region, also can help a firm decide where to locate a new plant, a warehouse, or a retail or service outlet. By comparing domestic production data from the censuses with information on the exports and imports of individual commodities, a firm can determine if it is taking full advantage of the sales potential overseas. On the other hand, if its domestic sales are not performing as well as in the past, this type of comparison may show increased import competition. Some individual firms look to their trade associations or private research firms for the summary data needed for such internal analysis. While these organizations are excellent sources of key figures about an industry’s performance, frequently the firm is unaware of the original source of much of these data. Trade associations republish Census Bureau data, at times modifying the data to meet the needs of a particular industry. Not infrequently, the associations or research firms benchmark their own data series to the census data or have special tabulations of the data made by the Census Bureau. Sometimes a failure to acknowledge the Census Bureau as the source of the data is because there is no legal requirement to cite data that are in the public domain. At other times, they are meticulous in citing the sources of information, but the harried business user overlooks these footnotes. These are just a few of the possible ways that the firm could use census data in its own operations. Many firms are making such analyses and have seen the possibility of having the Census Bureau further tabulate the data to shed more light on their own individual activities. Quite frequently, such firms request that the Census Bureau regroup establishment data, product data, or materials consumed data into geographic or other groupings that better fit their individual needs. Such tabulations are done when the firm making the request can pay the cost, provided that the results will not disclose the operation of individual companies. Such tabulations can lead to more intelligent marketing and production strategies.
TAKING ADVANTAGE OF THIS SNAPSHOT OF THE ECONOMY Many census data users-or, more accurately, many of those who could but do not take advantage of this wealth of data-fail to understand the basic role of the economic censuses as a snapshot of the economy at a point in time. Users frequently complain that the results are not timely and, therefore, are of little value. But timeliness is not the primary purpose of the censuses. And the results are never truly timely since they are taken at 5-year intervals; must be taken in the year after the period covered; and require 2 to 3 years to conduct, analyze, and publish. Timeliness is the primary objective of the current surveys, which usually are based on relatively small samples of establishments or companies in the target sector or industries. But because of their small size, the sample surveys provide only minimal detail. For extensive detail, the censuses are required.
258
GOVERNMENT
INFORMATION
QUARTERLY
Vol. ~/NO. 311987
So where does this leave the analyst, market researcher, or others, who must have both current and detailed statistics? Quite obviously it presents a problem but not an insoluble one. Faced with this common problem, the knowledgeable user of economic statistics consults the unique data that are available from the censuses for what they are, rather than rejecting them for what they are not. Because the data always will be from 2 to 7 years out of date, their value rests in (1) their use as benchmark measures, (2) the long-term trends they depict, and (3) the ratios or relationships between statistical measures that can be calculated. A short discussion of the use of each of these characteristics will suggest the virtually endless possibilities for such indirect use.
l
l
l
To estimate current sales, capital expenditures, or other measures available from the census for a specific industry, product, or geographic area, estimates of change since the census benchmark year can be applied to the census figure. Such estimates of change may be based on current surveys, trade association statistics, a company’s own operation or study, or whatever measure has been determined to be appropriate. Frequently, the appropriateness of the estimate of change that has been selected can be tested by applying comparable estimates for earlier periods to data for earlier census years, then comparing the results with the latest census enumeration. Current estimates frequently are derived by simply extrapolating long-term trends. Such estimates usually are modified by the use of current data that show the extent of deviation from such long-term trends. Or the more sophisticated user may apply complex correlation or regression statistical techniques to arrive at current estimates. In either case, the long-term trends data series is the starting point. The use of ratios or averages such as those described earlier is, perhaps, the most versatile of the ways the censuses can be used indirectly. Ratios tend to be much more stable over time than the actual data values. And this stability, or the trend in the change in the ratio, can be tested by computing comparable ratios over successive censuses. By using such ratios together with current estimates for one of the ratio’s components, current estimates for the other component are derived easily. For example, it may be determined that materials costs represented 45% of the value of shipments for a particular industry in the most recent census. The ratio of .45 may be applied to projected shipments for a future period to provide a crude estimate of materials costs for that period. Such estimating techniques are used routinely in all types of planning. But their application to census data frequently is overlooked.
There are, of course, a number of cautions to observe and pitfalls to avoid in such indirect but essential uses of the censuses. Some will be obvious; others must be learned from experience. For example, it is important when using ratios to determine that the relationship between components is not affected by different inflation factors or, if so, that an appropriate adjustment is made to compensate for such differences. In all such cases of using census data or other sources of data, personal experience and common
Use of Economic
Census
259
Data
sense must be used. And always remember that the resulting estimates, despite their census origin, are only estimates. Their accuracy ultimately will be tested by the benchmark data from the next censuses. The economic censuses provide a comprehensive and detailed snapshot of the economy once each 5 years. Because of these characteristics, along with the historical availability of consistent data, the economic censuses serve as the foundation for the comprehensive economic statistics programs of the Federal government. The data are used to address a virtually endless array of economic questions as well as provide the benchmarks, weights, and sampling frames for accurate current measures of economic activity. The economic censuses are critical to accurate information about the nation’s economy and to the millions of decision makers actively involved in the economy. The data collections should continue to be conducted every 5 years because of rapid changes taking place in the economy. Basic characteristics of providing detailed data by industry and geographic location must be maintained. And improved coverage of the service activities in the economy will make the censuses even more useful in the future. NOTES AND REFERENCES 1. 2. 3.
4.
5.
For a description of the Standard Industrial Classification System, see Lawrence A. Blum et al., “Conducting the Economic Censuses,” Government Information Quarterly. 4 (1987): 289-29 1. For a detailed discussion of how the census data are made available, see Paul T. Zeisset, “Making Economic Census Data Available,” Government Information Quarterly, 4( 1987): 303-324. Included in the monthly surveys are the Current Industrial Reports (this series also includes annual and quarterly surveys), Manufacturers’ Shipments, Inventories, and Orders, Monthly Retail Trade: Sales ana’ Inventories, and Monthly Wholesale Trade: Safes and Inventories. In the benchmarking process, sample survey estimates for the census year are revised to agree with the totals enumerated in the census. Survey estimates for subsequent periods similarly are adjusted in line with the benchmark period. For discussion of the benchmarking process, see Manfactarers’ Shipments, Inventories, and Orders: 1977-1982, M3-1.12, Current Industrial Reports (Washington, DC: Bureau of the Census, May 1983), p. 3. Also refer to Malcolm E. Bemhardt and Sol D. Helfand, “Reconciliation of Economic Censuses Results and Current Surveys Program,” Economic and Demographic Statistics, selected papers given at the 1980 Annual Meeting of the American Statistical Association (Washington, DC: Social Security Administration, November 1980), p. 4 1. Producer Price Index weights are revised when data from the quinquennial industrial censuses, as well as sufficient budgetary resources, become available. The weights represent the total net selling value of commodities produced, processed, or imported in this country and flowing into primary markets. Each census provides the latest comprehensive data on the net selling value for most of the commodities produced or processed. For further information, see “Producer Prices,” Bureau of Labor Statistics Handbook of Methods, Volume 1, Chapter 7, Bulletin 2134-I (Washington, DC: Bureau of Labor Statistics, 1982).
GAYLORD WORDEN
The comprehensive economic censuses conducted by the Bureau of the Census at 5-year intervals are the foundation of the nation’s economic statistics programs. These censuses play a critical role in allowing millions of private and public decision makers to make more informed economic plans and decisions. The uses of these data are limited only by the failures of human imagination. This article describes briefly some of the principal uses of the economic censuses. The United States economy is large and complex. Millions of private and public sector
decision makers are involved daily in keeping it healthy and vigorous. They rely on information from the nation’s economic statistics programs to better understand the complexities and reduce the uncertainties surrounding the decisions they make. The economic censuses, conducted once each 5 years, are the foundation of the nation’s economic statistics programs. They are the most comprehensive collection of data on the many sectors of the economy. Because these data are consistent, reliable, and detailed, they are used in a myriad of ways to address the structure and performance of the economy. They are used to address detailed questions on inputs used and outputs produced by industry or kind of business, both nationally and in small geographic areas. And they provide controls and sampling frames for surveys that produce more current but less detailed measures of economic activity. This article describes the major uses of data from the economic censuses. These uses are: l l l l l
framework for current economic surveys; measuring and tracking changes in economic activity; calculation of composite measures of the nation’s economic activity; analysis of economic policies; planning and monitoring of economic programs; and planning and management by private sector businesses.
Gaylord Worden is Chief; Industry Division, U.S. Bureau of the Census, Washington, D.C. 20233. Government Infommtion
Quarterly, Volume 4, Number 3 (1987), pages 245-259.
ISSN: 0740-624X.
246
GOVERNMENT
INFORMATION
Before these uses are discussed in some detail, a description content of the economic censuses is provided.
A DETAILED
SNAPSHOT
QUARTERLY
Vol. ~/NO. 311987
of the basic structure
and
OF THE ECONOMY
The economic censuses often are described as providing a detailed snapshot of the economy. This snapshot is taken for years ending in “2” and “7.” It is the comprehensive coverage of economic sectors, the industry or kind of business and geographic area detail, and the reliability and availability of the data that make the censuses so useful. No other statistical program provides similar data on economic activity on such a scale. Most sectors of the economy are included in the economic censuses. There are separate but coordinated censuses for manufacturing; mining; wholesale trade; retail trade; business, personal and professional services; and construction, Some transportation sector industries are covered while public utilities, finance, insurance, and real estate are the principal sectors not included. Regulatory programs traditionally have provided data for these sectors. Agriculture and governmental units are covered in separate but parallel censuses.
EstablishmentData For each of the economic sectors covered by the censuses, each plant, store, or other business location with employees is identified and included in the censuses. Each location is commonly called an establishment. The establishment may differ from the company or firm, which is an organization consisting of one or more establishments under common ownership or control. Nonemployer establishments are also included in the retail trade, services, and construction censuses. For each establishment, data are collected on employment, payroll, and other inputs and costs. Comparable data on shipments, receipts, and other measures of output or production help complete the snapshot of economic activity. The amount of detail on inputs and production varies significantly by sector. The most detail is provided for manufacturing where data on nearly 1,000 input materials and over 11,000 products produced are collected in the census. Data for other sectors generally are much less detailed. Based on data reported on the primary products produced at a manufacturing plant or the major kind of business conducted at a business location, the establishment is classified using the Standard Industrial Classification (SIC) system into the 4-digit industry.’ This system is used widely for statistical data by Federal agencies as well as the private sector to provide consistency in the way data are classified and to allow users to consult data from many sources. Data from the economic censuses provide detailed information by industry because they are classified at the 4-digit level. Each establishment also is assigned a detailed geographic code. This code allows it to be classified by city or place, county, metropolitan area, and state. Again, no other statistical programs provide such comprehensive and consistent economic data for these small geographic areas for the entire nation. Table 1 (pp. 248-249) shows the major data items collected in the censuses by geographic level of detail.
Use of Economic
Census
Data
247
Because data are collected by establishment and classified by a detailed industry description and geographic location, establishment data are often called building blocks for statistical information. These building blocks provide: comprehensive coverage of most of the economy; consistent basic input and output, or production data across the economy; l data for detailed industry classifications; l data for small geographic areas; l great flexibility in how the data can be aggregated and used; and * another data point each 5 years that is consistent with an already long series of data.
l
l
EnterpriseData An additional building block included in the censuses is enterprise-(company) level data. By assigning a common identifier to all establishments of the same company, the Census Bureau is able to put together all the establishments for companies, providing a rich set of data for analyzing structure and organization of economic activity. Making this linkage between information for enterprises and establishments available once each 5 years also allows data users to make better use of the variety and combinations of independent establishment and company-level data available in other years. For example, data used in calculating the Gross National Product (GNP) accounts quarterly and annually include establishment data from many surveys as well as statistical information from company-level reports filed with the Internal Revenue Service. Other Data In addition to the building blocks described above, other programs are included in the economic censuses to provide selected data once every 5 years. One sample survey collects inventory and use information on the nation’s fleet of trucks. A special program provides data on minority-owned business enterprises, and another provides data on women-owned businesses. Still others provide detailed data on water use by manufacturers and mining operations, and on the types of textile machinery being used. PRIMARY USES OF CENSUS DATA Only the failures of human imagination limit the uses made of data from the economic censuses. Because these data measure the structure and activity of a large portion of the economy on a consistent and scheduled basis, they constitute a major body of facts and information that are used at all levels of government and in the private sector. The primary and most frequent uses are described in this section. But because the ways the economic census data might be used are truly limitless, they are so widely available,2 and there are so many users, it is not possible to chronicle all users and uses.
.. .. .. .. .. .. . . . . 0
*
248
.
.
.
.
.
.
.
*
*
.
.
1
.
.
.
.
‘
*
*
*
.
*
*
*
.
.
.
.
.
*
.
.
.
.
.
.
.
*
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
=
.
.
-
.
.
.
.
.
... ... .. .. . . .*.
.. .. ..
.
.
.
*. .. .. 1 .
I
.
.
.
.
.
.
.
.
.
.
.
.
*
.
.
f
*
.
I
.
*
.
* *
._
.
.+a
.
. . f .
. , .
. * .
.
.*4J
. . it! . .k.6h D . “* . * * :zJa \it -1. 3 0. .l E :. .9-r :,oli 5 u: :trP p : :?:: Ia. * :2; ?I : :2,” $ : .*ai: I
.h .*, ..4
.e .C . .
r,
. .m .
.” . . .
.
.
.
.
. ..I-
:
SO .\
. .
.m . .
. :Jr; !z w. ‘“J
.a .o
.
.. ... .. . ... ..
... ... ... ... .
.. .. .. .. .. ..
.
.
.. .. .. .. ... .
.* ... .. .* ...
1
.
.
*
*
.
.
*
*
.
*
*
*
.
.
.
I
.
*
.
*
.
.
.
.
.
.
.
.
*
*
.
.
.
.
.
.
.
.
.
.
::
ii
.
* . . .
::
i5
:L :: .h . :B :%S . . :4 ;g ;g :;’ :=i
;
:;+” ** 0 *alo
*c
::
G’XSS
$2
249
.. .. .. i.. ...
2
250
GOVERNMENT
Framework
INFORMATION
QUARTERLY
Vol. ~/NO. 3/l 987
for Current Economic Surveys
Since the censuses are only conducted once each 5 years, it is necessary to have more frequent sample surveys to measure economic activity and to update the 5-year censuses. There are annual sample surveys of manufacturing, retail trade, wholesale trade and services. Important monthly surveys3 also are conducted for manufacturing and retail trade. The censuses provide the best available and most complete list of business establishments, and key information about their activities. The list is needed to draw the sample panels for a survey. Being able to draw a new sample once each 5 years from an up-to-date list helps assure the accuracy of these sample surveys. With the rapid changes taking place in the economy, less frequent development of new samples would lead to problems in the accuracy of these current measures of the economy. Some of the current sample surveys also are benchmarked to or reconciled with the censuses.4 Since the complete coverage in the censuses provides more accurate data than the surveys that are subject to sampling error, the benchmarking process improves the sample estimates. In a similar way, some trade associations benchmark data they collect to the censuses. Measuring and Tracking Changes in Economic Activity
The comprehensive and detailed data from the economic censuses are used by all types of users to address a myriad of questions about changes in economic activity. Because these building blocks of data can be put together in so many ways, they are highly useful for such analytical activities as charting long-term trends in the economy; identifying what industries and businesses exist, their relative importance, and their geographic location; and measuring what kinds of material they consume, how much labor they employ, and how much capital they require. A few examples of broad questions that can be answered are: l
l
l
l l l
l
l
What is the changing mix of manufacturing industries and which are growing or declining most rapidly? Are multiestablishment companies becoming more vertically integrated or horizontally diversified? Are department stores accounting for a smaller or larger share of total clothing sales? What has been the shift in textile employment by geographic regions (see Figure l)? Are small firms increasing their share of road building activity? Which states have the biggest increases in business services activity? Has the ranking of states by level of business services receipts changed in the past 10 years? Do minority-owned businesses account for a larger share of economic activity than they did 10 years ago? In which types of economic activity are minority-owned businesses increasing their shares most rapidly? In which states are Hispanicowned businesses increasing most rapidly? Have manufacturing sales branches and oftices increased their share of wholesale activities since 1972?
Figure 1.
100
Percent Textile Employment
in New England and South Atlantic
States.
Percent a m
New England South Atlantic
90
80
60
En-
30 24
20
10
1963 Year 251
1972
1962
Figure 2.
Share of Value of Shipments
Four and Eight Largest Companies Percent (Concentration
in the Electronic
Accounted Computing
for by the Equipment
Industry.
ratio)
100 m m
Four Largest Companies Eight Largest Companies
90 83 80
1967
1972
1977 Year 252
1982
Use of Economic
Census Data
253
Other questions may focus on a particular geographic area, kind of business, or product produced, or the relationship between two industries or two economic variables. A few examples are: l
l l
l
l
l
l
l
l
What changes are occurring in manufacturing employment in the Cleveland metropolitan area? Is employment growing in some primary metals industries and declining in others? Are food store receipts higher in Waterloo or Cedar Rapids, Iowa? Were receipts the same in 1977? How does the change in manufacturing activity in North Carolina compare with other southeastern states? What are some of the products in which aluminum is being used more? What materials is it replacing? Is the manufacturing of electronic computing equipment becoming more concentrated (see Figure 2)? Is retail activity in the center of Ft. Worth growing or declining relative to surrounding suburban areas? How has the mix of labor and material costs changed in the past 15 years for the manufacturers of dairy products? How specialized is the wood furniture industry?
These are a few examples of what can be an endless stream of very diverse questions. All potential users of the data, whether in a local, state, or Federal government organization or in a small or large business, ask these types of questions about economic activity. Only the economic censuses provide the broad but detailed data required to answer such a wide variety of questions. Some of the broader questions listed above also can be answered with data from annual or more frequent sample surveys. But such an ability to measure and track trends in economic activity with these more current surveys is dependent on the economic censuses, which provide the framework for the current surveys. Composite Measures of the Nation’s Economic Activity
Some of the best known economic statistics series are those that provide current composite measures of economic status and trends. These include quarterly estimates of GNP and nonfarm productivity, and monthly series such as the Index of Industrial Production and the Producer Price Index. The economic censuses are critical in maintaining accuracy of each of these. Benchmarking the national income and product accounts (NIPA), including the level of GNP and all its components, to data from the economic censuses is an extremely important process. Since no other data provide the detail and accuracy of the economic censuses, the GNP estimates, benchmarked to the censuses, are the most complete and accurate possible. This benchmarking is a very detailed and involved process carried out by the Bureau of Economic Analysis (BEA).
254
GOVERNMENT
INFORMATION
QUAKTERLY
Vol. ~/NO. 311987
The first step in this process is to use data from the censuses as well as many other sources to prepare a detailed picture of the economy showing the production and use of commodities and services among industries. This picture, known as input-output tables, is prepared by tracing how each of the economy’s industries use the products of all industries in producing final products and services for consumption or investment by other industries. This requires the detailed product and material data collected in the censuses as well as information on sales by class of customer which is collected in every other census or once each 10 years. These tables are useful for analyses of how changes in demand for consumer and investment goods, government expenditures, exports, and imports affect all industries. These updated input-output tables then are used to derive benchmark estimates of the NIPA. All of this is dependent on the data from the economic censuses. While the input-output tables use data from all the economic censuses, the manufacturing and mining industries comprise about 70% of the commodity classes covered by the tables. These complex tables are prepared only for economic census years when the detailed data are available. The Bureau of Labor Statistics (BLS) and the Bureau of Economic Analysis (BEA) also use data from the censuses to update the weights assigned to various products in important price series. The BLS uses data from the Census of Manufactures on the quantity of products shipped to determine how much weight to assign to the price of each of several thousand products included in the Producer Price Index in order to summarize them into more aggregate index numbers. In a similar manner, BEA relies on census data for its price series maintained for deflating GNP estimates. Updated weights from the censuses5 help maintain the accuracy of price series by reflecting the current mix of products being produced. Accuracy of price series is, in turn, vital for accurate estimates of real economic activity, that is, estimates adjusted to reflect changes in prices. A third major way census data are used as the foundation for current composite measures of economic activity is for benchmarking and weighting the monthly Index of Industrial Production by the Federal Reserve Board and the quarterly estimates of nonfarm productivity by BLS. Periodic updating of weights for these series are required because of the changing mix of industries over a 5-year period. And as this mix changes, the Ievels of these series can be corrected only by benchmarking to the economic censuses. Providing this foundation for the composite economic measures is one of the more important uses of the economic censuses data. Many decisions on the content and scope of coverage of the censuses are made based on how that decision will affect their use in composite measures. Having accurate composite measures is critical for the development of informed public economic policies and for private decision making. All investors, consumers, wage earners, and producers are affected by the accuracy of current composite measures and their contribution to reducing the uncertainties surrounding the decisions they have to make. Analysis of Economic Policies: Planning and Monitoring
of Economic Programs
Many economic policies and programs require more detailed information than what is available from composite measures of the national economy. Frequently the greater
2.55
Use of Economic Census Data
detail required is by small geographic area; sometimes it is by industry or product. The economic censuses serve these needs very well. A common use is for planning or monitoring economic development programs for a local area. Or the data may be used to assess the geographic distribution of economic activity or changes in that dist~bution to determine the need for regional economic assistance programs. Another important use is to develop plans and programs for the nation in the event of an emergency. Others want to evaluate the performance of programs to encourage the development of minority-owned businesses or small businesses. Another example is the use of concentration ratio data to judge the performance of existing antitrust programs or the need for additional programs. Only data available from the economic censuses once each 5 years can serve all these purposes. These are examples of policies and programs that are common to all levels of government. State and local bodies make frequent use of the data by small geographic area. The censuses give them a rich, though infrequent, source of data for decisions on how to provide efficient services and plan for economic growth. These same users in all levels of government rely on the less detailed data from annual and more frequent surveys for analysis of economic policies and programs. Since the accuracy of these more current data is vital for these uses, economic censuses also are critical in this indirect way. Planning and Management
by Private Sector Businesses
While the uses of census data already described are important to private sector business managers, we need also to describe and highlight some of the specific and direct ways the private sector can use census data. It is important to recognize that many of the suppliers of the data, the manufacturer, the retailer, or the service establishment also are among the primary users of the results. The more knowledgeable they are concerning how census data can benefit them, the more likely they are to participate in the censuses and surveys and to respond accurately and quickly. A major task facing any statistical agency is to educate its suppliers concerning how they can use the data. Since not all firms are aware of the direct benefits to be realized from census results, some examples of how they can use the data to improve both production and marketing decisions and strategy are useful. /~~roy~ng
P~ucfjon
Strategy
Census data can be used by individua1 firms in planning and developing their production strategy. In many cases, this strategy depends on assessing the performance of their firm against other firms in the same industry or geographic area. By comparing itself with those in the same general category, that is, in the same industry, same geographic area, same relative size class, or those specializing in the same types of products, the firm compares its own performance to the industry as a whole. For example, using the latest data from the Census of Manufactures or the Annual Survey of Manufactures series, the firm can compare its performance to certain industry averages. Averages or ratios that may be used for comparison include the following: l
payroll per dollar of sales;
256
GOVERNMENT
INFORMATION
QUARTERLY
Vol. ~/NO. 3/1987
l cost of materials per dollar of shipments; 0 payroll per employee; l annual wages per production worker; l production worker wages per hour; l nonproduction workers to production workers; l capital expenditure per employee or dollar sales; and l unit value (value/quantity) for individual products.
The various measures that could be considered for judging the performance of the individual operating unit can be extended to as many data items as are collected. As part of the Census of Manufactures, the reports provide many ratios for each industry over a number of years in order to make firm/industry comparisons easy. By comparing changes from year to year for products, industries, or geographic areas, with changes that have taken place in the firm itself, the company can determine whether it is improving its performance relative to the rest of the industry. If its performance is unduly different, the firm can investigate the possible causes. Using the census data, the firm can compare itself to other firms classified in its industry in terms of the mix of specific materials consumed or the mix of labor and materials’ costs. If its mix is noticeably different, and the firm has been losing its market share, it may want to investigate why. Other comparisons useful in production planning can be made from census data. For example, a table from the Census of Manufactures breaks down the value of industry shipments according to the primary and secondary products that are typically produced in the industry. Some of this industry product mix evolves naturally as a by-product of the manufacturing process. Secondary production may be a logical extension of using some of the waste from an earlier stage of production. In other cases, secondary products have evolved because individual firms have found it advantageous to produce commodities that can be marketed together with their primary product. The individual firm can examine how its product mix compares with other firms in the industry and examine whether there may be a more effective product mix. Improving
Market Strategy
Census data also permit the individual firm to evaluate its position in a market or determine the existence of a potential market. For example, the individual firm could analyze its share of the production of an individual product in the United States or in a specific state. Or the firm may look at its share of a broad class of products produced in a particular state. If this information is too aggregated, individual firms frequently request that the Census Bureau provide special tabulations of individual products produced or individual materials consumed by manufacturing establishments according to geographic areas that fit their marketing or sales district. If the company produces a product that is used in other manufacturing processes, it can determine the industries that are most likely to use its product, and by studying the geographic location of those industries, identify the most fruitful marketing areas. If the firm is producing a product that is used by or distributed through other economic sectors, such as retailing, wholesaling, or construction, it can use data from
Use of Economic
Census Data
257
the economic censuses for those sectors to identify the best potential markets among these types of businesses. Similarly, a firm can use this information to lay out sales territories and to pinpoint markets for advertising purposes. Information on market potential and potential suppliers, if used in conjunction with knowledge of the demographic characteristics of a region, also can help a firm decide where to locate a new plant, a warehouse, or a retail or service outlet. By comparing domestic production data from the censuses with information on the exports and imports of individual commodities, a firm can determine if it is taking full advantage of the sales potential overseas. On the other hand, if its domestic sales are not performing as well as in the past, this type of comparison may show increased import competition. Some individual firms look to their trade associations or private research firms for the summary data needed for such internal analysis. While these organizations are excellent sources of key figures about an industry’s performance, frequently the firm is unaware of the original source of much of these data. Trade associations republish Census Bureau data, at times modifying the data to meet the needs of a particular industry. Not infrequently, the associations or research firms benchmark their own data series to the census data or have special tabulations of the data made by the Census Bureau. Sometimes a failure to acknowledge the Census Bureau as the source of the data is because there is no legal requirement to cite data that are in the public domain. At other times, they are meticulous in citing the sources of information, but the harried business user overlooks these footnotes. These are just a few of the possible ways that the firm could use census data in its own operations. Many firms are making such analyses and have seen the possibility of having the Census Bureau further tabulate the data to shed more light on their own individual activities. Quite frequently, such firms request that the Census Bureau regroup establishment data, product data, or materials consumed data into geographic or other groupings that better fit their individual needs. Such tabulations are done when the firm making the request can pay the cost, provided that the results will not disclose the operation of individual companies. Such tabulations can lead to more intelligent marketing and production strategies.
TAKING ADVANTAGE OF THIS SNAPSHOT OF THE ECONOMY Many census data users-or, more accurately, many of those who could but do not take advantage of this wealth of data-fail to understand the basic role of the economic censuses as a snapshot of the economy at a point in time. Users frequently complain that the results are not timely and, therefore, are of little value. But timeliness is not the primary purpose of the censuses. And the results are never truly timely since they are taken at 5-year intervals; must be taken in the year after the period covered; and require 2 to 3 years to conduct, analyze, and publish. Timeliness is the primary objective of the current surveys, which usually are based on relatively small samples of establishments or companies in the target sector or industries. But because of their small size, the sample surveys provide only minimal detail. For extensive detail, the censuses are required.
258
GOVERNMENT
INFORMATION
QUARTERLY
Vol. ~/NO. 311987
So where does this leave the analyst, market researcher, or others, who must have both current and detailed statistics? Quite obviously it presents a problem but not an insoluble one. Faced with this common problem, the knowledgeable user of economic statistics consults the unique data that are available from the censuses for what they are, rather than rejecting them for what they are not. Because the data always will be from 2 to 7 years out of date, their value rests in (1) their use as benchmark measures, (2) the long-term trends they depict, and (3) the ratios or relationships between statistical measures that can be calculated. A short discussion of the use of each of these characteristics will suggest the virtually endless possibilities for such indirect use.
l
l
l
To estimate current sales, capital expenditures, or other measures available from the census for a specific industry, product, or geographic area, estimates of change since the census benchmark year can be applied to the census figure. Such estimates of change may be based on current surveys, trade association statistics, a company’s own operation or study, or whatever measure has been determined to be appropriate. Frequently, the appropriateness of the estimate of change that has been selected can be tested by applying comparable estimates for earlier periods to data for earlier census years, then comparing the results with the latest census enumeration. Current estimates frequently are derived by simply extrapolating long-term trends. Such estimates usually are modified by the use of current data that show the extent of deviation from such long-term trends. Or the more sophisticated user may apply complex correlation or regression statistical techniques to arrive at current estimates. In either case, the long-term trends data series is the starting point. The use of ratios or averages such as those described earlier is, perhaps, the most versatile of the ways the censuses can be used indirectly. Ratios tend to be much more stable over time than the actual data values. And this stability, or the trend in the change in the ratio, can be tested by computing comparable ratios over successive censuses. By using such ratios together with current estimates for one of the ratio’s components, current estimates for the other component are derived easily. For example, it may be determined that materials costs represented 45% of the value of shipments for a particular industry in the most recent census. The ratio of .45 may be applied to projected shipments for a future period to provide a crude estimate of materials costs for that period. Such estimating techniques are used routinely in all types of planning. But their application to census data frequently is overlooked.
There are, of course, a number of cautions to observe and pitfalls to avoid in such indirect but essential uses of the censuses. Some will be obvious; others must be learned from experience. For example, it is important when using ratios to determine that the relationship between components is not affected by different inflation factors or, if so, that an appropriate adjustment is made to compensate for such differences. In all such cases of using census data or other sources of data, personal experience and common
Use of Economic
Census
259
Data
sense must be used. And always remember that the resulting estimates, despite their census origin, are only estimates. Their accuracy ultimately will be tested by the benchmark data from the next censuses. The economic censuses provide a comprehensive and detailed snapshot of the economy once each 5 years. Because of these characteristics, along with the historical availability of consistent data, the economic censuses serve as the foundation for the comprehensive economic statistics programs of the Federal government. The data are used to address a virtually endless array of economic questions as well as provide the benchmarks, weights, and sampling frames for accurate current measures of economic activity. The economic censuses are critical to accurate information about the nation’s economy and to the millions of decision makers actively involved in the economy. The data collections should continue to be conducted every 5 years because of rapid changes taking place in the economy. Basic characteristics of providing detailed data by industry and geographic location must be maintained. And improved coverage of the service activities in the economy will make the censuses even more useful in the future. NOTES AND REFERENCES 1. 2. 3.
4.
5.
For a description of the Standard Industrial Classification System, see Lawrence A. Blum et al., “Conducting the Economic Censuses,” Government Information Quarterly. 4 (1987): 289-29 1. For a detailed discussion of how the census data are made available, see Paul T. Zeisset, “Making Economic Census Data Available,” Government Information Quarterly, 4( 1987): 303-324. Included in the monthly surveys are the Current Industrial Reports (this series also includes annual and quarterly surveys), Manufacturers’ Shipments, Inventories, and Orders, Monthly Retail Trade: Sales ana’ Inventories, and Monthly Wholesale Trade: Safes and Inventories. In the benchmarking process, sample survey estimates for the census year are revised to agree with the totals enumerated in the census. Survey estimates for subsequent periods similarly are adjusted in line with the benchmark period. For discussion of the benchmarking process, see Manfactarers’ Shipments, Inventories, and Orders: 1977-1982, M3-1.12, Current Industrial Reports (Washington, DC: Bureau of the Census, May 1983), p. 3. Also refer to Malcolm E. Bemhardt and Sol D. Helfand, “Reconciliation of Economic Censuses Results and Current Surveys Program,” Economic and Demographic Statistics, selected papers given at the 1980 Annual Meeting of the American Statistical Association (Washington, DC: Social Security Administration, November 1980), p. 4 1. Producer Price Index weights are revised when data from the quinquennial industrial censuses, as well as sufficient budgetary resources, become available. The weights represent the total net selling value of commodities produced, processed, or imported in this country and flowing into primary markets. Each census provides the latest comprehensive data on the net selling value for most of the commodities produced or processed. For further information, see “Producer Prices,” Bureau of Labor Statistics Handbook of Methods, Volume 1, Chapter 7, Bulletin 2134-I (Washington, DC: Bureau of Labor Statistics, 1982).