Technical note.

Brief Explanation of Producer Price Indexes

The Producer Price Index (PPI) of the Bureau of Labor Statistics (BLS) is a family of indexes that measure the average change over time in the prices received by domestic producers of goods and services. PPIs measure price change from the perspective

of the seller. This contrasts with other measures, such as the Consumer Price Index (CPI). CPIs measure price change from the purchaser's perspective. Sellers' and purchasers' prices can differ due to government subsidies, sales and excise taxes, and distribution costs.

More than 8,000 PPIs for individual products and groups of products are released each month. PPIs are available for the products of virtually every industry in the mining and manufacturing sectors of the U.S. economy. New PPIs are gradually being introduced for the products of industries in the construction, trade, finance, and services sectors of the economy.

More than 100,000 price quotations per month are organized into three sets of PPIs: (1) Stage-of processing indexes, (2) commodity indexes, and (3) indexes for the net output of industries and their products. The stage-of-processing structure organizes products by class of buyer and degree of fabrication. The commodity structure organizes products by similarity of end use or material composition. The entire output of various industries is sampled to derive price indexes for the net output of industries and their products.

Stage-of-Processing Indexes

Within the stage-of-processing system, finished goods are commodities that will not undergo further processing and are ready for sale to the final-demand user, either an individual consumer or business firm. Consumer foods include unprocessed foods such as eggs and fresh vegetables, as well as processed foods such as bakery products and meats. Other finished consumer goods include durable goods such as automobiles, household furniture, and appliances, as well as nondurable goods such as apparel and home heating oil. Capital equipment includes durable goods such as heavy motor trucks, tractors, and machine tools.

The stage-of processing category for intermediate materials, supplies, and components consists partly of commodities that have been processed but require further processing. Examples of such semifinished goods include flour, cotton yarn, steel mill products, and lumber. The intermediate goods category also encompasses nondurable, physically complete items purchased by business firms as inputs for their operations. Examples include diesel fuel, belts and belting, paper boxes, and fertilizers.

Crude materials for further processing are products entering the market for the first time that have not been manufactured or fabricated and that are not sold directly to consumers. Crude foodstuffs and feedstuffs include items such as grains and livestock. Examples of crude nonfood materials include raw cotton, crude petroleum, coal, hides and skins, and iron and steel scrap.

The illustration above shows examples of how products are classified by stage of processing.

Commodity Indexes

The commodity classification structure of the PPI organizes products by similarity of end use or material composition, disregarding industry of origin. Fifteen major commodity groupings (two-digit commodity codes) make up the All Commodities Index. Each major commodity grouping includes (in descending order of aggregation) subgroups (three-digit codes), product classes (four-digit codes), subproduct classes (six-digit codes), and individual items (eight-digit codes). Nearly all eight-digit commodities under the traditional commodity coding system are now derived from corresponding industry-classified product indexes. In such instances, movements in the traditional commodity price indexes and corresponding percent changes will be virtually identical to their industry-based counterparts, even if their index levels differ.

Industry Net-Output Price Indexes

PPIs for the net output of industries and their products are grouped according to the North American Industry Classification System (NAICS). Prior to the release of January 2004, industry-based PPIs were published according to the Standard Industrial Classification (SIC) system. Industry price indexes are compatible with other economic time series organized by industry, such as data on employment, wages, and productivity. Table 5 of the PPI Detailed Report includes data for NAICS industries and industry groups (3-, 4-, 5-, and 6-digit codes); Census product classes (7- and 8-digit codes), products (9-digit codes), and more detailed subproducts (11-digit codes); and, for some industries, indexes for other sources of revenue.

Indexes may represent one of three kinds of product categories. Every industry has primary product indexes to show changes in prices received by establishments classified in the industry for products made primarily, but not necessarily exclusively, by that industry. The industry classification of an establishment is determined by which products make up a plurality of its total shipment value. In addition, most industries have secondary product indexes that show changes in prices received by establishments classified in the industry for products chiefly made in some other industry. Finally, some industries have miscellaneous receipts indexes to show price changes in other sources of revenue received by establishments within the industry that are not derived from sales of their products--for example, resales of purchased materials, or revenues from parking lots owned by a manufacturing plant.

Data Collection

PPIs are based on selling prices reported by establishments of all sizes selected by probability sampling, with the probability of selection proportionate to size. Individual items and transaction terms from these firms also are chosen by probability proportionate to size. BLS strongly encourages cooperating companies to supply actual transaction prices at the time of shipment to minimize the use of list prices. Prices submitted by survey respondents are effective on the Tuesday of the week containing the 13th day of the month. This survey is conducted primarily through the mail.

Price data are provided on a voluntary and confidential basis; only sworn BLS employees are allowed access to individual company price reports. BLS publishes price indexes instead of actual prices. All PPIs are subject to revision 4 months after original publication to reflect the availability of late reports and corrections by respondents.

BLS periodically updates the PPI sample of survey respondents to better reflect current conditions when the structure, membership, technology, or product mix of an industry shifts significantly and to spread reporting burden among smaller firms. Results of these resampling efforts are incorporated into the PPI with the release of data for January and July.

As part of an ongoing effort to expand coverage to sectors of the economy other than mining and manufacturing, an increasing number of service sector industries have been introduced into the PPI. The table on the following page lists recently introduced industries and includes the month and year in which an article describing the industry's content appeared in the PPI Detailed Report.

Weights

Weights for most traditional commodity groupings of the PPI, as well as weights for commodity-based aggregate indexes calculated using traditional commodity groupings, such as stage-of-processing indexes, currently reflect 1997 values of shipments as reported in the Census of Manufactures and other sources. From January 1996 through December 2001, PPI weights were derived from i992 shipment values. Industry indexes also are now calculated with i997 net output weights. This periodic update of the value weights used to calculate the PPI is done to more accurately reflect changes in production and marketing patterns in the economy. Net output values of shipments are used as weights for industry indexes. Net output values refer to the value of shipments from establishments within the industry to buyers outside the industry. However, weights for commodity price indexes are based on gross shipment values, including values of shipments between establishments within the same industry. As a result, broad commodity grouping indexes, such as the PPI for All Commodities, are affected by the multiple counting of price change at successive stages of processing, which can lead to exaggerated or misleading signals about inflation. Stage-of-processing indexes partially correct for this defect, but industry indexes consistently correct for it at all levels of aggregation. Therefore, industry and stage-of-processing indexes are more appropriate than are broad commodity groupings for economic analysis of general price trends.

Price Index Reference Base

Effective with publication of January. 1988 data, many important PPI series (including stage-of-processing groupings and most commodity groups and individual items) were placed on a new reference base, 1982 = 100. From 1971 through 1987, the standard reference base for most PPI series was 1967 = 100. Except for rounding differences, the shill to the new reference base did not alter any previously published percent changes for affected PPI series. (See "Calculating Index Changes," below.) The 1982 reference base is not used for commodity indexes with a base later than December 1981 or for industry net output indexes and their products.

For further information on the underlying concepts and methodology of the Producer Price Index, see chapter 14, "Producer Prices," in BLS Handbook of Methods (April 1997), Bulletin 2490. This document can be downloaded from the BLS Web site at www.bls.gov/opub/hom/homch14_itc.htm Reprints are available on request.

                                                    PPI Detailed

                 Title                     Code      Report Issue

                                          SIC

Wireless telecommunications               4812        July 1999
Telephone communications,
  except radio telephone                  4813        July 1995
Television broadcasting                   4833        July 2002
Grocery stores                            5411        July 2000
Meat and fish (seafood) markets           5421        July 2000
Fruit and vegetable markets               5431        July 2000
Candy, nut, and confectionery stores      5441        July 2000
Retail bakeries                           5461        July 2000
Miscellaneous food stores                 5499        July 2000
New car dealers                           5511        July 2000
Gasoline service stations                 5541     January 2002
Boat dealers                              5551     January 2002
Recreational vehicle dealers              5561     January 2002
Miscellaneous retail                      59       January 2001
Security brokers, dealers, and
  investment bankers                      6211     January 2001
Investment advice                         6282     January 2003
Life insurance carriers                   6311     January 1999
Property and casualty insurance           6331        July 1998
Insurance agencies and brokerages         6412     January 2003
Operators and lessors of
  nonresidential buildings                6512     January 1996
Real estate agents and managers           6531     January 1996
Prepackaged software                      7372     January 1998
Data processing services                  7374     January 2002
Home health care services                 8082     January 1997
Legal services                            8111     January 1997
Engineering design, analysis, and
  consulting services                     8711     January 1997
Architectural design, analysis, and
  consulting services                     8712     January 1997
Premiums for property and casualty
  insurance                               9331        July 1998

                                          NAICS

New warehouse building construction       236221      July 2005
Merchant wholesalers, durable goods       423         July 2005
Merchant wholesalers, nondurable
  goods                                   424         July 2005
Wholesale trade agents and brokers        425120      July 2005
Furniture and home furnishings stores     442      January 2004
Electronics and appliance stores          443      January 2004
Building material and garden
  equipment and supplies dealers          444      January 2004
Clothing and clothing accessories
  stores                                  448      January 2004
Sporting goods, hobby, book, and music
  stores                                  451      January 2004
General merchandise stores                452      January 2004
Miscellaneous store retailers             453      January 2004
Internet service providers                518111      July 2005
Web search portals                        518112      July 2005
Commercial banking                        522110   January 2005
Savings institutions                      522120   January 2005
Direct health and medical insurance
  carriers                                524114      July 2004
Construction, mining, and forestry
  machinery and equipment rental
  and leasing                             532412   January 2005
Security guards and patrol services       561612      July 2005
Fitness and recreational sports centers   713940      July 2005

Calculating Index Changes

Each PPI measures price changes from a reference period that equals 100.0. An increase of 5.5 percent from the reference period in the Finished Goods Price Index, for example, is shown as 105.5. This change also can be expressed in dollars, as follows: prices received by domestic producers of a sample of finished goods have risen from $100 in 1982 to $105.50. Likewise, a current index of 90.0 would indicate that prices received by producers of finished goods are 10 percent lower than they were in 1982.

Movements of price indexes from one month to another are usually expressed as percent changes, rather than as changes in index points. Index point changes are affected by the level of the index in relation to its base period, whereas percent changes are not. The following example shows the computation of index point and percent changes.

Index point change

Finished Goods Price Index             107.5
Less previous index                    104.0
Equals index point change                3.5

Index percent change

Index point change                       3.5
Divided by the previous index          104.0
Equals                                 0.034
Result multiplied by 100         0.034 x 100
Equals percent change                    3.4

Seasonally Adjusted and Unadjusted Data

Because price data are used for different purposes by different groups, BLS publishes seasonally adjusted and unadjusted changes each month. Seasonally adjusted data are preferred for analyzing general price trends in the economy because these data eliminate the effect of changes that normally occur at about the same time, and in about the same magnitude, every year--such as price movements resulting from normal weather patterns, regular production and marketing cycles, model changeovers, seasonal discounts, and holidays. For these reasons, seasonally adjusted data more clearly reveal underlying cyclical trends. Unadjusted data are of primary interest to users who need information that can be related to actual dollar values of transactions. Individuals requiring this information include marketing specialists, purchasing agents, budget and cost analysts, contract specialists, and commodity traders. It is the unadjusted data that are generally cited when escalating long-term contracts such as purchasing agreements or real estate leases. For more information, see Escalation and Producer Price Indexes: A Guide for Contracting Parties, BLS Report 807, September 1991, on the Web at www.bls.gov/ppi/ ppiescalation.htm. Reprints are available on request.

In 1998, the PPI implemented the X-12-ARIMA Seasonal Adjustment Method; prior to that year, the PPI employed the X-11 method. Each year, the seasonal status of most commodity indexes is reevaluated to reflect more recent price behavior. Industry net output indexes are not seasonally adjusted. For time series that exhibit seasonal pricing patterns, new seasonal factors are estimated and applied to the unadjusted data for the previous 5 years. These updated seasonally adjusted indexes replace the most recent 5 years of seasonal data.

Seasonal factors may be applied to series using either a direct or an aggregative method. Generally, commodity indexes are seasonally adjusted using direct seasonal adjustment, which produces a more complete elimination of seasonal movements than does the aggregative method. However, the direct seasonal adjustment process may not yield figures that possess additive consistency. Thus, a seasonally adjusted index for a broad category that is directly adjusted may not be logically consistent with all seasonally adjusted indexes for its components. Seasonal movements for stage-of-processing indexes are derived indirectly through an aggregative method that combines movements of a wide variety of subproduct class (six-digit) series.

Seasonally adjusted indexes can become problematic when previously stable and predictable price patterns abruptly change. If the new pattern persists, the seasonal adjustment method will eventually reflect it adequately; if the pattern keeps shifting, however, seasonally adjusted data will become chronically troublesome. This problem occurs infrequently for farm and food-related products, but more often for manufactured products such as automobiles and steel.

Since January 1988, the PPI has used Intervention Analysis Seasonal Adjustment methods to enhance the calculation of seasonal factors. With this technique, outlier values that may distort the seasonal pattern are removed from the data prior to applying the standard seasonal factor estimation procedure. For example, a possible economic cause for large price movements for petroleum-based products might have been the Persian Gulf War. In this case, intervention techniques allowed for better estimates of seasonally adjusted data. On the whole, very few series have required intervention. Out of nearly 900 seasonally adjusted series, only 16 were subject to intervention in 1997.

For more information relating to seasonal adjustment methods, see (1) "Appendix A: Seasonal Adjustment Methodology at BLS," in the BLS Handbook of Methods (April 1997), Bulletin 2490 and (2) "Summary of Changes to the PPI's Seasonal Adjustment Methodology" in the January 1995 issue of Producer Price Indexes.

Producer Price Index Data on the Internet

In 1995, the BLS began posting PPI series, news releases, and technical information to both a World Wide Web (WWW) site and a file transfer protocol (FTP) site. During the years following the introduction of PPI Internet services, use of these sites eclipsed more traditional methods of data dissemination, such as subscriptions to the PPI Detailed Report. There were more than 1.6 million instances of PPI series being downloaded from the Internet during the 12 months ended December 31,2003.

Retrieving PPI data from the PPI Web site

PPI data can be obtained from the WWW address (www.bls.gov/ppi). Scrolling down the page to the "Get Detailed PPI Statistics" header reveals the following methods of data retrieval:

* Most Requested Series is a form-based application that allows the user to quickly obtain PPI time series data by selecting from two separate lists (commodity and industry) of the most commonly requested time series, including the All Commodities Index and the stage-of-processing indexes (for example, Finished Goods). Within each list, any one--or all--of the time series shown can be selected. A user can modify the date range and output options after executing the query, using the reformat button above the data output table.

* Create Customized Tables is a form-based query application designed for users unfamiliar with the PPI coding structure. The application guides a user through the PPI classification system by listing index titles and does not require knowledge of commodity or industry codes. Data retrieved are based on a query formulated by selecting data characteristics from lists provided. Two options are available to create customized tables, depending on a user's browser capability. The one-screen option is a JavaScript application that uses a single screen to guide a user through the available time series data. The second option is a multiple-screen, non-Java-based application. Both methods allow a user to browse the PPI coding structure and select multiple series codes. Using the one-screen option, users can modify the date range and output options after executing the query using the reformat button above the data output table.

* Series Report is a form-based application that uses formatted PPI time series identifiers (commodity or industry codes) as input in extracting data according to a specified set of date ranges and output options. This application provides the most efficient path for users who are familiar with the format of PPI time series identifiers. Up to 300 indexes can be extracted at a time.

There are five alphabetic prefixes used to create unique PPI time series identifiers: WP, WD, PC, PD, and ND. Each provides the user access to a different PPI database. Adding either a "u" (not seasonally adjusted) or an "s" (seasonally adjusted) to the end of these prefixes further specifies the type of data needed.

EXAMPLES

For commodity and stage-of-processing indexes, series identifiers combine a "wpu" prefix (not seasonally adjusted) or a "wps" prefix (seasonally adjusted) with a commodity code.

Commodity code    Provides data for

wps141101         Passenger cars, seasonally
                    adjusted
wpu141101         Passenger cars, not
                    seasonally adjusted
wpusop3000        Finished goods, not
                    seasonally adjusted

For discontinued commodity indexes, series identifiers combine a "wdu" prefix (not seasonally adjusted) or a "wds" prefix (seasonally adjusted) with a commodity code.

Commodity code    Provides data for

wds019            Other farm products, seasonally
                    adjusted
wdu0635           Preparations, ethical (prescription),
                    not seasonally adjusted
wdusi138011       Stainless steel mill products,
                    not seasonally adjusted

Current price indexes grouped by industry according to NAICS have series identifiers that begin with the prefix "pcu." After the prefix, there are 12 digits (the 6-digit industry code is listed twice) followed by up to 7 alphanumeric characters identifying product detail. Dashes are used as placeholders for higher-level industry group codes.

Industry-product code,
current NAICS series      Provides data for

pcu325---325---           Chemical manufacturing, not
                            seasonally adjusted
pcu336110336110           Automobile and light duty
                            motor vehicle manufacturing
pcu621111621111411        Offices of physicians, one- and
                            two-physician practices and
                            single-specialty group practices,
                            general/family practice
pcu325412325412A          Pharmaceutical preparation
                            manufacturing, pharmaceuticals
                            acting on the respiratory
                            system

Discontinued industry-product codes based on SIC combine a "pdu" prefix and "#" between the fourth and fifth characters of the product code. Series identifiers for the discontinued dataset use underscores as placeholders to complete a reference to an SIC industry group code of fewer than four digits. (All PPI industry-based indexes organized by SIC were discontinued with the introduction of NAICS.)

Industry-product code,
discontinued SIC series    Provides data for

pdu28_ _#                  Chemicals and allied products,
                             not seasonally adjusted
pdu331_#                   Blast furnaces, steel works,
                             and rolling and finishing
                             mills, not seasonally adjusted
pdu3711#111                Passenger cars

Price indexes for discontinued series grouped by industry according to NAICS have identifiers that begin with the prefix "ndu." After the prefix, there are 12 numeric digits (the 6-digit industry code is listed twice), and up to 7 additional alphanumeric characters that identify product detail. Dashes are used as placeholders for higher-level industry group codes.

Industry-product code,
discontinued NAICS
series                    Provides data for

ndu212231212231           Lead ore and zinc ore mining
ndu2122312122312          Lead and zinc concentrates
ndu212231212231214        Lead concentrates

* Flat Files and the FTP server are best suited for users requiring access to either a large volume of time series data or other PPI-related documentation (such as seasonal factor and relative importance tables). The FTP site can be accessed at ftp://ftp.bls.gov or directly from the links on the "Get Detailed Statistics" page or the PPI homepage.

Data and documentation available for download include:

                             Directory

NAICS current series         /pub/time.series/pc
SIC discontinued series      /pub/time.series/pd
NAICS discontinued series    /pub/time.series/nd
Commodity data,
  current series             /pub/time.series/wp
Commodity data,
  discontinued series        /pub/time.serieslwd
Special requests             /pub/special.requests/ppi
Latest news release          /pub/news.release/ppi.txt

The FTP site maintains files to help with searches and downloads. These files are centrally located in the/pub/doc directory. Within this directory, the overview.txt file contains an overview relating to all BLS data available through the FTP site For current commodity-based PPI data, the program help file is wp.txt; for discontinued commodity series, wd.txt; for current industry-based PPI data based on NAICS, pc.txt; for industry-based SIC time series that have been discontinued, pd.txt; and for industry-based NAICS series, nd.txt.

Users who prefer downloading PP! datasets as individual ZIP files should go to the directory labeled/pub/time.series/ compressed/tape.format/on the FTP site. This directory includes six PPI-specific ZIP files, one for each of the PPI databases--WP, WD, PC, ND, and PD--and a ZIP file for the annual 5-year revision to historical seasonal PPIs.

Other Sources of PPI Data

PPI data can also be accessed via the BLS homepage (www.bls.gov). Clicking on the "Get Detailed Statistics" link at the top of the homepage calls up a chart listing all available BLS programs. The following methods are available for retrieving PPI data: Most requested statistics, create customized tables (one screen or multiple screens), and flat files. Additional sources of BLS data also are accessible from this page, including economic news releases, series report, and economy at a glance.

Additional information

The PPI homepage (www.bls.gov/ppi) contains additional information regarding PPI data and methodology. The top section of the homepage provides PPI news releases, both current and archived, as well as general PPI information. The "Tables Created by BLS" section found beneath the statistics section provides relative importance and seasonal factor tables. The remaining sections offer special notices and publications pertaining to PPI methodology and applications.

For questions or comments regarding PPI data classification, methodology, or data availability on the Internet, call or e-mail the Section of Index Analysis and Public Information at (202) 691-7705 or ppi-info@bls.gov.