new weight structure for producer price index - fraser · 2018. 11. 7. · ^classifications which...

9
New Weight Structure for Producer Price Index Andrew Clem and William D. Thomas For the first time in over 10 years, the universe of weights used to construct the Producer Price Index (PPI) has been comprehensively updated. The dynamic nature of the Ameri- can economy requires that periodic revisions be made to the PPI weight structure to ensure its relevance. The revision affects all PPI series derived from traditional commodity indexes, including stage-of-processing indexes, durability- of-product indexes, and special commodity groupings. Historical indexes are not affected, and there is no break in continuity of the PPI. Weights are used to aggregate price changes for individual commodities to higher level groupings. They represent the total net selling value of commodities priced in the PPI, which comprises a wide range of crude and intermediate materials and finished goods, including both capital equip- ment and goods intended for household consumption. Between January 1976 and December 1986, the data used to derive the weights were taken from the 1972 values of shipments. Since January 1987, 1982 values have been used for weighting purposes. This article describes the procedures for updating the weights, presents the revised and former relative importance values, and discusses some of the causes of major shifts in weights that resulted from the revision. Derivation of indexes and weights Most of the basic data on selling values of commodities in the PPI are taken from the economic censuses of the Bureau of the Census. The 1982 Census of Manufactures provided the data for about 84 percent of the total value; the 1982 Census of Mineral Industries (which includes oil and gas production) accounted for about 7 percent, and the 1982 Census of Wholesale Trade (covering scrap materials) for 0.6 percent. Other sources were the Department of Agricul- ture (5.1 percent), the Edison Electric Institute (3.4 percent), and the National Marine Fisheries Service (0.1 percent). The formula by which the PPI is calculated is a modifica- tion of the Laspeyres fixed-base-weight formula. In the pure Laspeyres formula, the physical quantities of all individual items are held constant, regardless of price change.* Although Andrew Clem and William D. Thomas are economists in the Office of Prices and Living Conditions, Bureau of Labor Statistics. this assumption is suitable for medium-term time spans, structural shifts among industries necessitate occasional weight revisions. In practice, however, the PPI is computed using total values of shipments rather than physical (or unit) quantities, which are not available on a consistent and com- prehensive basis, as are values of shipments. There is an im- plicit fixed quantity associated with each specific commodity (bushels, tons, gallons, etc.), but it is never computed. 2 The alternative to the Laspeyres formula would be the Paasche formula, which uses current-quantity weights instead of base-quantity weights. However, this method has its own drawbacks and, in any case, is impractical for an index com- bining price movements for thousands of items every month. The equations on page 5 describing the weight-update procedure are followed by an explanation of their use. (Be- cause published index and relative importance values have a reference base of 100, each of the formulas is multiplied by 100.) In the following equations: P = price (dollars per unit for commodities; can only be defined for groupings as the ratio of the total values between two time periods) Q = quantity (number of physical units) V = value (of shipments); or P times Q R = relative importance (the ratio of the aggregate value of an item or mid-level grouping to the total, e.g., all commodities, value) I = index 67 = total or average for the year 1967 72 = total or average for the year 1972 82 = total or average for the year 1982 86 = December 1986 (the weight-link month) b = base year (in general) t = current month (in general) i = individual commodity A = aggregate-level index (e.g., all commodities) N = number of commodities within the aggregate 1 The assumption of fixed quantities imparts to the Laspeyres formula a tendency to overstate the importance of items whose prices have risen sharply since the weight base-period, and to understate the importance of items that have declined in price. See, for example, William Mendenhall and James E. Reinmuth, Statistics for Management and Economics, Duxbury Press 1974, p. 421. 2 Base-year quantities could be computed if base-year prices were known. However, the Bureau of Labor Statistics does not maintain files of prices on a historical basis. Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis April 1987

Upload: others

Post on 05-Oct-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

New Weight Structurefor Producer Price Index

Andrew Clem and William D. Thomas

For the first time in over 10 years, the universe of weightsused to construct the Producer Price Index (PPI) has beencomprehensively updated. The dynamic nature of the Ameri-can economy requires that periodic revisions be made to thePPI weight structure to ensure its relevance. The revisionaffects all PPI series derived from traditional commodityindexes, including stage-of-processing indexes, durability-of-product indexes, and special commodity groupings.Historical indexes are not affected, and there is no break incontinuity of the PPI.

Weights are used to aggregate price changes for individualcommodities to higher level groupings. They represent thetotal net selling value of commodities priced in the PPI,which comprises a wide range of crude and intermediatematerials and finished goods, including both capital equip-ment and goods intended for household consumption.

Between January 1976 and December 1986, the data usedto derive the weights were taken from the 1972 values ofshipments. Since January 1987, 1982 values have been usedfor weighting purposes. This article describes the proceduresfor updating the weights, presents the revised and formerrelative importance values, and discusses some of the causesof major shifts in weights that resulted from the revision.

Derivation of indexes and weights

Most of the basic data on selling values of commoditiesin the PPI are taken from the economic censuses of theBureau of the Census. The 1982 Census of Manufacturesprovided the data for about 84 percent of the total value; the1982 Census of Mineral Industries (which includes oil andgas production) accounted for about 7 percent, and the 1982Census of Wholesale Trade (covering scrap materials) for0.6 percent. Other sources were the Department of Agricul-ture (5.1 percent), the Edison Electric Institute (3.4 percent),and the National Marine Fisheries Service (0.1 percent).

The formula by which the PPI is calculated is a modifica-tion of the Laspeyres fixed-base-weight formula. In the pureLaspeyres formula, the physical quantities of all individualitems are held constant, regardless of price change.* Although

Andrew Clem and William D. Thomas are economists in the Office of Pricesand Living Conditions, Bureau of Labor Statistics.

this assumption is suitable for medium-term time spans,structural shifts among industries necessitate occasionalweight revisions. In practice, however, the PPI is computedusing total values of shipments rather than physical (or unit)quantities, which are not available on a consistent and com-prehensive basis, as are values of shipments. There is an im-plicit fixed quantity associated with each specific commodity(bushels, tons, gallons, etc.), but it is never computed.2

The alternative to the Laspeyres formula would be thePaasche formula, which uses current-quantity weights insteadof base-quantity weights. However, this method has its owndrawbacks and, in any case, is impractical for an index com-bining price movements for thousands of items every month.

The equations on page 5 describing the weight-updateprocedure are followed by an explanation of their use. (Be-cause published index and relative importance values havea reference base of 100, each of the formulas is multipliedby 100.)

In the following equations:

P = price (dollars per unit for commodities; can only be definedfor groupings as the ratio of the total values between two timeperiods)

Q = quantity (number of physical units)V = value (of shipments); or P times QR = relative importance (the ratio of the aggregate value of an item

or mid-level grouping to the total, e.g., all commodities, value)I = index

67 = total or average for the year 196772 = total or average for the year 197282 = total or average for the year 198286 = December 1986 (the weight-link month)

b = base year (in general)t = current month (in general)i = individual commodity

A = aggregate-level index (e.g., all commodities)N = number of commodities within the aggregate

1 The assumption of fixed quantities imparts to the Laspeyres formulaa tendency to overstate the importance of items whose prices have risensharply since the weight base-period, and to understate the importance ofitems that have declined in price. See, for example, William Mendenhalland James E. Reinmuth, Statistics for Management and Economics,Duxbury Press 1974, p. 421.

2 Base-year quantities could be computed if base-year prices were known.However, the Bureau of Labor Statistics does not maintain files of priceson a historical basis.

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 2: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

(1) N

I pi,.Qi,b

lA,t

1 pi,bQi,bi = 1

(2)

r , (Pi,t /Pi,67) (Pi,67Qi,82)xA,t =

4,82

(3)

N,86 (72) =

Vi,72 (Pi,86

VA,72 (PA,86 / P A , 7 2 )

_ Pi,72 Qi,72 (Pi,86 /Pi,72)

Pi,72 Qi,72 ( £ Pi,86 Qi,72 / £ Pi,72 Qi.72)

i = 1

P i ,86 Qi,72

N

1 Pi,86Qi,72

i = 1

i = 1

N

1i = 1

(4)

R i ,86 (82)

A ,82

Pi,82 Qi,82 (Pi,86 /Pi,82)

N N

1 Pi,82Qi,82(li = 1 i = 1

Pi,86 Qi,82

Pi,82Qi,82)

i = 1

Pi,86 Qi,82

i = 1

A number of complicating factors necessitate modificationsof the pure fixed-base-weight Laspeyres formula shown inequation (1). First, the arithmetic reference base year hasgenerally differed from the weight base period in the PPLAlso, as noted above, it is the aggregate values of commodi-ties, rather than their physical quantities, that are used tocalculate commodity grouping indexes. Aggregate (i.e., theabsolute weights) values are the prices multiplied by quanti-ties summed for each commodity in the grouping, andupdated for price change from the weight base year to theindex calculation month. The modified Laspeyres formulawhich approximates the calculation of the PPI is shown inequation (2).

More significantly, for the last several years there havebeen a number of sample expansions as well as semiannual^classifications which have required minor adjustments tothe weight structure. Relatively minor changes will continueas industry samples are "recycled" to account for structuralchanges in the U.S. economy. However, to facilitate theexplanation of the weight revision procedures, it is assumedthat the PPI base-year price measurements coincide perfectlywith the real market, and that no sample changes haveoccurred.

Equations (3) and (4) show how the relative importanceof each individual item is derived, taking base-year quanti-ties and adjusting them by PPI price movements between theweight base year (1972 or 1982) and December 1986.3 Therelative importance of an item will be higher (or lower) thanits share of the actual census total value of shipments if itsprice rose more (or less) during that time interval than didthe price level of the aggregate category to which it is beingcompared. For example, petroleum products have a muchlower weight than they would have if their prices had notfallen between 1982 and 1986.

Since both the old and revised weights have been adjustedfor price change through December 1986, any shifts observedmust derive from changes in quantities.

Causes of shifts in weights

Table C shows the relative importance of commoditygroupings in each stage-of-processing (SOP) category inDecember 1986 (the month in which the new and old weightswere linked) under the 1982 and 1972 weight structures. Thedifference between the two columns reflects primarily theshift in value weights. To a small extent, it also reflects theroutine sample changes that went into effect in January 1987.

One cause of the shifts in weights was a change in thedefinition of the PPI universe; the PPI now includes militarysales and also shipments between establishments of the samecompany (interplant transfers). Interplant transfers are nowincluded because, under the industry-oriented concept of thePPI revision, the universe of transactions includes all ship-ments outside the industry where a product originates,

3 The weight of an item or grouping is the absolute dollar value adjustedfor price change. The relative importance is that weight divided by the to-tal index weight and multiplied by 100.

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 3: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

Table C. Changes in relative importance of Producer Price Index commodity groupings by stage of processing,resulting from update of weight base from 1972 to 1982

Comtn-odi ty !code I

01011012013014015016017018019

02022301025

04041

050510531056

06062065202

07071103

08085

090912

101011011101210210211023

131321139

0202102202302402502530254026027028029

03031032033034038039

04042043044

05

Index

CRUDE MATERIALS FOR FURTHER PROCESSING

Farm productsFresh and dried fruits and vegetablesGralnsLivestockLive poultryPlant and animal fibersFluid milkEggsHay> hayseeds> and oilseedsOther farm products

Processed foods and feedsUnprocessed fin fishSugar and confectionery

Hides, skins, leather, and related productsHides and skins

Fuels and related products and powerCoalNatural gasCrude petroleum (domestic production)

Chemicals and allied productsPaints and allied productsPhosphates

Rubber and plastic productsReclaimed rubber

Lumber and wood productsLogst bolts> timber and pulpwood

Pulp, paper, and allied productsNastepaper

Metals and metal productsIron and steel

Iron oreIron and steel scrap

Nonferrous metalsNonferrous metal oresNonferrous scrap

Nonmetallic mineral productsConstruction sand, gravely and crushed stoneOther nonmetallic minerals

INTERMEDIATE MATERIALS, SUPPLIES, AND COMPONENTS

Processed foods and feedsCereal and bakery productsMeats, poultry, and fishDairy productsProcessed fruits and vegetablesSugar and confectionery

Refined sugar, Confectionery materials

Beverages and beverage materialsFats and oilsMiscellaneous processed foodsPrepared animal feeds

Textile products and apparelSynthetic fibersProcessed yarns and threadsGray fabricsFinished fabricsApparel and other fabricated textile productsTextile fibers, yarns and fabrics, n.e.c.

Hides, skins, leather, and related productsLeatherFootwearOther leather and related products

Fuels and related products and power

Relative iDecember

Revised(1982)

100.000

43.6021.4885.53119.4923.6020.8947.2670.2863.7431.299

0.8330.3330.500

0.6130.613

40.85910.73312.09618.030

0.3220.0560.266

0.0270.027

2.8992.899

1.1481.148

6.8533.3820.5892.7933.4711 .4442.027

2.8472.2160.631

100.000

5.1390.3710.9780.5130.1130.6440.3380.2560.3070.3850.2391.589

4.7000.6500.9890.9751.2840.6530.149

0.2750.2130.0040.058

12.128

mportance* 1986 1/

Former(1972)

100.000

60.5192.0206.79928.7083.3271.344

11.3490.4004.5282.044

1.4060.2981.108

0.8530.853

22.9526.0189.7437.191

0.3960.1020.294

0.0720.072

2.4252.425

0.4960.496

5.8303.0710.1452.9262.7590.4932.266

5.0513.8221.229

100.000

5.0910.3661.0600.4650.1160.8990.6770.1720.2740.3350.1961.380

5.0420.7500.9800.9451.5940.6750.098

0.3520.2540.0070.091

11.450

Ratio 2/

(1982/1972)

1.00000

0.720470.736630.813500.678971.082660.665180.640320.715000.826630.63552

0.592461.117450.45126

0.718640.71864

1.780191.783481.241512.50730

0.813130.549020.90476

0.375000.37500

1.195461.19546

2.314522.31452

1.175471.101274.062070.954551.258062.929010.89453

0.563650.579800.51343

1.00000

1.009431.013660.922641.103230.974140.716350.499261.488371.120441.149251.219391.15145

0.932170.866671.009181.031750.805520.967411.52041

0.781250.838580.571430.63736

1.05921

See footnotes at end of table.

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 4: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

Table C. Changes in relative importance of Producer Price Index commodity groupings by stage of processing,resulting from update of weight base year from 1972 to 1982—Continued

Comm-oditycode

0520532054057058

06061062063064065066067

07071072

0808108110812082083084086087

09091n<M xU71 d

09140915092093

10101102103104105106107108

11111112113114114311481161171178118119

12121122123124125126

131 3 1 I1322133134135136137138139

14141142143

Index

Coke oven productsLiquefied petroleum gasElectric powerPetroleum products, refinedPetroleum and coal productsy n.e.c.

Chemicals and allied productsIndustrial chemicalsPaints and allied productsDrugs and PharmaceuticalsFats and oils, inedibleAgricultural chemicals and chemical productPlastic resins and materialsOther chemicals and allied products

Rubber and plastic productsRubber and rubber productsPlastic products

Lumber and wood productsLumberSoftwood lumberHardwood lumber

MillworkPlywoodOther wood productsPrefabricated wood buildings and componentsTreated wood and contract wood preserving

Pulpy papery and allied productsPulpy papery and products> ex. building paperPaperPaperboardConverted paper and paperboard products

Building paper and building board mill productsPublicationsy printed mattery and printing mat.

Metals and metal productsIron and steelNonferrous metalsMetal containersHardwarePlumbing fixtures and brass fittingsHeating equipmentFabricated structural metal productsMiscellaneous metal products

Machinery and equipmentAgricultural machinery and equipmentConstruction machinery and equipmentMetalworking machinery and equipmentGeneral purpose machinery and equipment

Fluid power equipmentAir conditioning and refrigeration equipment

Special industry machinery and equipmentElectrical machinery and equipmentElectronic components and accessories

Miscellaneous instrumentsMiscellaneous machinery

Furniture and household durablesHousehold furnitureCommercial furnitureFloor coveringsHousehold appliancesHome electronic equipmentOther household durable goods

Nonmetallic mineral productsGlassCementConcrete productsClay construction products except refractoriesRefractoriesAsphalt felts and coatingsGypsum productsGlass containersOther nonmetallic minerals

Transportation equipmentMotor vehicles and equipmentAircraft and aircraft equipmentShips and boats

Relative importanceDecember 1986 1/

Revised(1982)

0.1170.3667.2054.3190.121

11.1494.1151.2790.9340.0751.1761.8391.731

4.4721.4413.031

2.9290.9580.6050.3530.9330.4040.3490.1440.141

11.6697.0102 316i!on3.1880.1824.477

19.1795.4734.2981.2470.7330.2520.3233.2313.622

15.7790.2850.2391.1503.5580.4140.8770.3376.8022.8120.6162.792

0.8970.0530.1490.2680.2040.0610.162

4.3010.4570.3491.3130.1320.1330.2420.1600.5680.947

5.4503.7141.4790.135

Former(1972)

0.0950.3176.2944.6040.140

9.7213.2611.4010.6240.1211.2931.3731.648

4.0261.8202.206

3.7941.4671.1330.3341.0860.5640.3280.2520.097

11.2796.227

o!7873.3670.2584.794

23.2128.2094.3471.2680.9810.3700.4343.5734.030

12.4150.2490.2231.2863.5040.2571.1070.3134.3391.2660.4212.080

0.9780.0690.1590.2460.2430.0390.222

5.8970.4960.5271.8540.2260.2020.2900.1920.7161.394

4.9473.7340.8220.082

Ratio 2/

(1982/1972)

1.231581.154571.144740.938100.86429

1.146901.261880.912921.496790.619830.909511.339401.05036

1.110780.791761.37398

0.772010.653030.533981.056890.859120.716311.064020.571431.45361

1.034581.12574A /LA r\s.a1 • *rl UHO

1.284630.946840.705430.93388

0.826250.666710.988730.983440.747200.681080.744240.904280.89876

1.270961.144581.071750.894251.015411.610890.792231.076681.567642.221171.463181.34231

0.917180.768120.937111.089430.839511.564100.72973

0.729350.921370.662240.708200.584070.658420.834480.833330.793300.67934

1.101680.994641.799271.64634

See footnotes at end of table.

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 5: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

Table C. Changes in relative importance of Producer Price Index commodity groupings by stage of processing,resulting from update of weight base year from 1972 to 1982—Continued

Comm-oditycode

144149

15151152153154156157159

01011016017019

02021022022102220223023024025026027028029

030320330340380381010381020381030382

04043044

050510530570571057302058

0606106206306350636065067

07071072

08081082083

09091092093

10101

Index

Railroad equipmentTransportation equipment, n.e.c.

Miscellaneous productsToys* sporting goods, small arms, etc.Tobacco products> including stemmed and redriedNotionsPhotographic equipment and suppliesMedical, surgical, and personal aid devicesIndustrial safety equipmentOther miscellaneous products

FINISHED GOODS

Farm productsFresh and dried fruits and vegetablesFluid milk

Other farm products

Processed foods and feedsCereal and bakery productsMeats, poultry, and fishMeatsProcessed poultryUnprocessed and packaged fish

Dairy productsProcessed fruits and vegetablesSugar and confectioneryBeverages and beverage materialsFats and oilsMiscellaneous processed foodsPrepared animal feeds

Textile products and apparelProcessed yarns and threadsGray fabricsFinished fabricsApparel and other fabricated textile productsHomen's apparelMen's and boys' apparelGirls', children's, and infants' apparelTex t ile house furn i sh i ngs

Hides, skins, leather, and related productsFootwearOther leather and related products

Fuels and related products and powerCoalGas fuelsPetroleum products, refinedGasolineFuel oil *2

Petroleum and coal products, n.e.c.

Chemicals and allied productsIndustrial chemicalsPaints and allied productsDrugs and PharmaceuticalsEthical preparations (prescription)Proprietary preparations (over-counter)

Agricultural chemicals and chemical productOther chemicals and allied products

Rubber and plastic productsRubber and rubber productsPlastic products

Lumber and wood productsLumberMi 11workPlywood

Pulp, paper> and allied productsPulp, paper> and products, ex. building paperBuilding paper and building board mill productsPublications* printed matter* and printing mat.

Metals and metal productsIron and steel

Relative importanceDecember

Revised(1982)

0.0990.023

1.9110.0800.2460.0610.3020.5290.0950.598

100.000

1.7251.2260.0740.3920.033

27.5743.3337.0464.8271.2400.9793.7701.7421.4455.6740.5143.3880.662

7.5000.0350.0530.1297.2833.3892.3870.5440.691

0.9900.6900.300

8.6310.0023.7774.7183.5860.5800.134

4.8160.0230.0321.9171.4300.4230.1062.738

1.4620.5560.906

0.1010.0240.0670.010

3.5611.3550.0212.185

1.0360.002

' 1986 1/

Former(1972)

0.2600.049

1.7970.0730.1000.1550.3140.2580.0630.834

100.000

1.8141.2900.0810.4250.018

26.4723.7247.1495.4180.7690.9623.6341.7741.2495.1130.4262.9520.451

6.8900.0380.0530.1426.6572.4312.6110.4680.850

1.3630.9850.378

7.4700.0022.4644.8503.5770.6320.154

4.5330.0130.0341.7331.2130.4780.1242.629

1.3960.8160.580

0.1160.0350.0680.013

3.5611.1270.0122.422

1.1820.003

Ratio 2/

(1982/1972)

0.380770.46939

1.063441.095892.460000.393550.961782.050391.507940.71703

1.00000

0.950940.950390.913580.922351.83333

1.041630.895010.985590.890921.612481.017671.037420.981961.156931.109721.206571.147701.46785

1.088530.921051.000000.908451.094041.394080.914211.162390.81294

0.726340.700510.79365

1.155421.000001.532870.972781.002520.917720.87013

1.062431.769230.941181.106171.178900.884940.854841.04146

1.047280.681371.56207

0.870690.685710.985290.76923

1.000001.202311.750000.90215

0.876480.66667

See footnotes at end of table.

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 6: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

Table C. Changes in relative importance of Producer Price Index commodity groupings by stage of processing,resulting from update of weight base from 1972 to 1982— Continued

Comm-oditycode

102103104106107108

111111121131137113811391141161161116211651167117117611811911911193

12121122123124125126

13131133138139

14141141101141105141106142143144149

15151152153154155156157159

Index

Nonferrous metalsMetal containersHardwareHeating equipmentFabricated structural metal productsMiscellaneous metal products

Machinery and equipmentAgricultural machinery and equipmentConstruction machinery and equipmentMetalworking machinery and equipmentMetal cutting machine toolsMetal forming machine toolsTools, dies> jigs, fixtures, and indust. molds

General purpose machinery and equipmentSpecial industry machinery and equipment

Food products machineryTextile machinery and equipmentPrinting trades machinery and equipmentPacking and packaging machinery

Electrical machinery and eauipmentCommunication and related equipment

Miscellaneous instrumentsMiscellaneous machineryOil field and gas field machineryOffice and store machines and equipment

Furniture and household durablesHousehold furnitureCommercial furnitureFloor coveringsHousehold appliancesHome electronic equipmentOther household durable goods

Nonmetallic mineral productsGlassConcrete productsGlass containersOther nonmetallic minerals

Transportation equipmentMotor vehicles ana equipment

Passenger carsLight trucksHeavy trucks

Aircraft and aircraft equipmentShips and boatsRailroad equipmentTransportation equipment, n.e.c.

Miscellaneous productsToys> sporting goods, small arms, etc.Tobacco products> including stemmed and redriedNotionsPhotographic equipment and suppliesMobile homesMedical, surgical, and personal aid devices IIndustrial safety equipmentOther miscellaneous products

Relative l mportanceDecember 1986 1/

Revised(1982)

0.0190.0040.1390.0490.4680.355

15.3321.1950.8931.7590.4410.1390.5521.9591.8940.1780.0840.2890.1065.0012.2460.9371.6940.4510.739

6.1711.6891.0610.5501.4070.5440.920

0.1130.0370.0260.0080.042

13.94410.2616.4402.2210.3081.7981.5700.2180.097

7.0451.139 I2.3330.0191.1220.484 j0.601 I0.009 !1.338 !

Former(1972)

0.0200.0060.1760.0520.5570.368

14.0061.3611.5101.8370.4140.2230.6531.8272.2090.1100.1460.2170.1852.3490.4090.5152.3980.1761.549

6.6752.1570.9270.6121.6360.3730.970

0.1700.0400.0210.0090.100

16.63913.5888.8112.0991.1951.3291.0040.5180.200

7.7071.1462.4820.0500.7230.9870.3490.0081.962

Ratio 2/

(1982/1972)

0.950000.666670.789770.942310.840220.96467

1.094670.878030.591390.957541.065220.623320.845331.072250.857401.618180.575341.331800.572972.128995.491441.819420.706422.562500.47708

0.924490.783031.144550.898690.860021.458450.94845

0.664710.925001.238100.888890.42000

0.838030.755150.730901.058120.257741.352901.563750.420850.48500

0.914100.993890.939970.380001.551870.490371.722061.125000.68196

1/ Relative importance data for commodity groupings include only those subproduct classesallocated to the respective stage-of-processmg (SOP) grouping. The revised weight structure(based on 1982 shipment values) includes the effects of sample changes in January 1987, while theold weight structure does not include these sample changes. Because these figures are based onunrevised December 1986 index levels, they are subject to revision) final relative importancedata will be published in the annual supplement to Producer Price Indexes later this year.

2/ A ratio above 1.0 indicates an increase in relative importance) for example, a ratio of1.25000 indicates a 25-percent increase, a ratio of 3.10000 indicates a 210-percent increase,and a ratio of 0.67000 indicates a 33-percent decrease.

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 7: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

regardless of ownership of the producing establishment.4 Theinclusion of military sales has had a minimal impact on stage-of-processing relative importance values because such salesdo not fit into any of the SOP categories.

Interplant transfers were previously excluded because thetraditional PPI was a measure of price change in primarymarkets, i.e., at the first commercial transaction for eachcommodity. For example, a steel manufacturer might owna coal mine supplying a blast furnace it operates at a differentlocation. The shipment of coal from the mine to the blastfurnace would not be a market transaction because the respec-tive establishments are part of the same corporate enterprise.Under the industry-oriented concept of the revised PPImethodology, all shipments outside the industry where aproduct originates are included. In this example, the coalmine and the blast furnace are classified as being in differ-ent industries even though they are owned by the sameenterprise. Thus, transactions between such commonly-owned establishments are included in the new PPI.

Of greater significance than changes in the definition ofthe universe were changes reflecting cyclical and long-termtrends in the economy. The economy of 1972 was stronglybolstered by stimulative monetary policies, while prices formost goods were frozen by Administration order. However,the Arab oil embargo of 1973 caused an oubreak of inflationthat did not abate until after economic stabilization meas-ures were taken that brought on a period of severe recessionduring the early 1980's.

As a result, the economy displayed unusual weakness inmany sectors during 1982, when most recent economiccensuses were taken, particularly in such industries as hous-ing and motor vehicles. Because of the sharp business cyclecontrast between 1972 and 1982, a number of categories inthe PPI related to these sectors showed substantial declinesin relative importance, unrelated to any long-term structur-al change in the economy.

The more significant weighting shifts that occurred are dis-cussed in the following paragraphs and shown in table C,which presents the relative importance values of commoditygroupings for each respective stage-of-processing categorybefore and after the weight revision.5 Commodity groupingsare listed in code-number sequence within each SOP group,similar to table 2 of the monthly report, Producer PriceIndexes.

Crude materials

The change in the scope of transactions included in thePPI universe was most pronounced within the crude materialscategory. All three types of crude energy materials—coal,natural gas, and crude petroleum—rose sharply because of

4 The PPI Revision (PPIR) program began in 1978 and was completed in1986. Virtually all 493 mining and manufacturing industries are now co-vered. However, the PPIR industry-classified indexes were not affected bythe weight revision; they continue to be calculated using 1977 net outputweights.

the inclusion of interplant transfers. The largest gain was forcrude petroleum, which rose 151 percent. This increaserepresented the inclusion of the output of oil wells ownedby vertically integrated petroleum refining companies. Therelative importance of coal climbed by 78 percent whilenatural gas rose about 24 percent. As a result, the total weightof energy in the crude materials category nearly doubled,from 23 percent on the 1972 basis to 41 percent on the 1982basis.

This unusually sharp increase for energy commoditiesresulted in a displacement of the importance of most othergoods; nonenergy crude goods overall showed a 23.2-percentdrop in relative importance. Thus, many items showed largeapparent decreases in market size when in fact there wereno major corresponding structural changes in the market.At the same time, this displacement phenomenon enhancesthe significance of the few increases which did occur fornonenergy items within the crude materials category, forexample, wastepaper, logs, and metal ores.

Among crude agricultural products, the largest decline inweight was for manufacturing grade raw milk (in the fluidmilk group), which fell 59 percent. Raw cane sugar declined45 percent in relative importance. The worldwide shortageof cane sugar that led to an explosion in prices during 1980prompted many confectionery and beverage producers toseek alternative ingredients such as artificial sweeteners. Theincreased popularity of diet beverages cut further into sugar'straditional markets.

Wheat registered a 37-percent drop in relative importance,due mostly to the large decrease in export demand in the wakeof the strengthening U.S. dollar during the early 1980's. Cornand soybeans showed much smaller declines, as these cropsare less dependent on export markets. The weights of bothcattle and hogs fell 32 percent, reflecting decreased consumerpreference for red meats in favor of poultry and fish; theselatter categories showed moderate increases.

Among crude materials other than energy and agriculturalproducts, the weak construction market in 1982 resulted insizable drops for construction sand, gravel, and crushed stoneand for miscellaneous nonmetallic minerals such as clay andmica. However, logs and timber showed a 20-percentincrease because of the inclusion of the values of shipmentsof timber to sawmills owned by the same company; this wasparadoxical because, as noted below, 1982 was a bad yearfor construction-related industries such as logging and saw-mills. The relative importance of wastepaper more thandoubled as environmental and economic considerationsencouraged paperboard manufacturers to make greater useof recycling as an alternative to reliance on woodpulp.

Metal ores rose very sharply, largely reflecting the inclu-sion of interplant transfers in the PPI weight universe. Amajority of metal mines in the U.S. are owned by the firms

5 Most of the data used in the analysis of long-term trends was takenfrom the 1984 U.S. Industrial Outlook, published by the U.S. Departmentof Commerce. This issue contained output data for many industries for both1972 and 1982.

10

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 8: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

which smelt or refine the ores themselves. Consequently,the weight of nonferrous ores nearly tripled, and the weightof iron ore more than quadrupled.

Intermediate goods

The weight shifts within the intermediate goods categoryappear consistent with the widespread perception thatAmerican industry has expanded in the high-technology andchemical-related sectors and declined in basic heavy indus-tries. However, the significant declines which occurred inthe relative importance of construction materials, such aslumber and nonmetallic minerals, were not due to such afundamental structural change in the economy, but rather tothe cyclical weakness of the construction sector in 1982.Neither the energy nor the food materials categories withinthe intermediate goods index showed much change in theweight revision.

The outstanding increase in relative importance among in-termediate goods was for electronic components and acces-sories, which more than doubled, from 1.3 percent to 2.8percent. The biggest gains were for integrated circuits, ofwhich some types increased tenfold or twentyfold in rela-tive importance between 1972 and 1982. Demand for someof these semiconductor chips was boosted enormously by themicrocomputer boom that began in the late 1970's; the in-crease in household sound and video equipment and com-mercial communication equipment sectors also stimulatedrapid growth in semiconductors.

Growth in the aerospace industry also was reflected in thePPI weight revision, although part of the increase stemmedfrom the inclusion of military sales in the weight universe.The relative importance of aircraft engines and engine partsnearly tripled, while aircraft parts and auxiliary equipmentrose 44 percent.

Medical and surgical instruments and appliances alsoshowed substantial increases in their weights, as shipmentsgrew in tandem with the increased proportion of medical ex-penses in household and public budgets. Growth was alsoevident among pharmaceuticals. Biological products morethan doubled their relative importance, and medicinal andbotanical chemicals (from which drugs are derived) showedan 83-percent advance in importance.

The chemical industry was given a tremendous stimulusby growth in plastic product markets. During the 1970's and1980's, new types of plastic materials came to be used moreand more frequently in durable goods. Plastic products usedby businesses recorded a 37-percent increase in relativeimportance, while plastic resins and materials rose 34 per-cent. Because of demand derived from the plastics sector andthe rapid rate of price increase since 1972, the relativeimportance of industrial chemicals moved up from 3.3 per-cent to 4.1 percent of the total intermediate goods index, a25-percent increase.

Another area of notable increases was the pulp and paperindustry. The relative importance of paper rose 41 percent,

while that of paperboard moved up 28 percent. Althoughfaced with competition from plastics and other packagingmaterials, the paper industry managed to develop new typesof containers made from corrugated paper.

An interesting substitution effect occurred in the processedfoods and feeds category. Confectionery materials gained 49percent in weight, reflecting increased use of corn sweetenerssuch as high-fructose corn syrup in candy and soft drinks.Corn sweeteners displaced refined sugar, which fell 50 per-cent and paralleled the drop in the relative importance of rawcane sugar within the crude materials index. In addition,crude vegetable oils rose 36 percent, while prepared animalfeeds showed a more modest increase, 15 percent.

Among industrial goods exhibiting major declines inrelative importance were asbestos products, which fell 61percent. Health concerns during the 1970's resulted in sharprestrictions on use of asbestos as an insulation material. Knitfinished fabrics declined 53 percent, and narrow-wovenfabrics moved down 44 percent.

Real total expenditures on residential construction in 1982fell to their lowest level since the early 1960's. Consequently,the weights assigned to most construction materials withinthe intermediate goods category dropped substantially; forexample, softwood lumber moved down from 1.1 percentto 0.6 percent. In addition, the weight of plywood fell near-ly 30 percent, and millwork registered a decrease of about 14percent. Clay bricks and tiles were down nearly 50 percent,and concrete products declined about 30 percent.

Basic metal industries also showed significant weightdeclines as domestic output of raw steel in 1982 dropped tothe lowest level since 1958. Besides the reduced demand forsteel resulting from the downsizing of automobiles and fewerpurchases of heavy machinery by American industries, asmuch as 20 percent of the diminished steel market was takenby imported steel. Thus, the relative importance of both steelmill products and foundry and forge shop products fellroughly one-third following the weight revision.

There was no clear trend among nonferrous metals, whichshowed virtually no net change overall. Aluminum gainedsubstantially in importance because of increased use in bever-age containers and aerospace equipment. However, zinc fellby about two-thirds, and precious metals declined 18 percent.

Contrary to the pattern of petroleum-derived and phar-maceutical chemicals, certain chemical products moved downin importance. Inedible fats and oils, mixed fertilizers, andmiscellaneous paint products all fell at least 30 percent.Various electrical equipment categories also fell (between10 and 25 percent), including electric lamps and bulbs, air-conditioning and refrigeration equipment, motors and gener-ators, and wiring devices.

The relative weight of the glass containers categorydecreased by about 20 percent as the increased popularityof aluminum, plastic, and paper containers (and stricter legis-lation on deposits for beverage containers in some States)displaced much of the market once held by the glass con-tainer industry.

11

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987

Page 9: New Weight Structure for Producer Price Index - FRASER · 2018. 11. 7. · ^classifications which have required minor adjustments to the weight structure. Relatively minor changes

Finished goods

Some of the sharpest changes in weights among finishedgoods took place in capital equipment. The weight of com-munication and related equipment nearly quintupled, reflect-ing the growing importance and increased applications of thistechnology. Among the most prominent examples of theseare the entry into the market of new telephone service com-panies and new television enterprises, particularly cable TV.Oilfield and gasfield machinery almost tripled in importance,as deregulation of domestic oil and the partial decontrol ofgas prices precipitated a high level of exploration and drill-ing activity and in turn sharp increases in demand for suchmachinery. Significant increases also occurred in the weightsfor photographic equipment, ships, and aircraft.

On the other hand, the relative importance of heavy trucksin the finished goods category dropped from 1.2 percent toonly 0.3 percent, reflecting the poor state of demand forautomotive products during the recession of 1981-82. Inaddition, the following types of capital goods declined inimportance: Locomotive parts (79 percent); textile machineryand equipment (42 percent); and railroad cars and parts (40percent). Because the Bureau now assigns the weight of thecomputer industry to the broad category of machinery andequipment, rather than to the more specific category of officeand store machinery, the weight of the latter dropped sub-stantially, 52 percent. In addition, the weights of metal-forming machine tools, truck trailers, plastics machinery,and automotive maintenance equipment also moved down.

Among consumer finished goods, one of the largest shiftswas the 27-percent decline in relative importance forpassenger cars, which now represent 6.4 percent of thefinished goods index, compared to 8.8 percent under theprevious weight structure. Sales of domestic cars fell to under6 million in 1982, compared to over 9 million in 1972.During the same period, sales of imported cars climbedsteadily. The growing consumer preference for light-dutypickup trucks was reflected in the 6-percent increase for lighttrucks. Given the nearly disastrous conditions in the motorvehicle market in 1982, the fact that the relative importancefor light trucks increased at all is significant.

Several other durable consumer goods declined in impor-tance from 1972 to 1982. The largest decline was for traveltrailers and campers, which fell 65 percent, mostly inresponse to curtailed vacation driving following the gasolineprice increases of the 1970's. Household furniture fell 22percent, jewelry and jewelry products declined 52 percent,and mobile homes decreased 50 percent over the same period.

One of the largest gains among consumer goods was forhome electronic equipment, where an increase of 93 percentwas noted for high-fidelity components and speakers. Sharpincreases were also registered for phonograph records andprerecorded tapes (62 percent) and television sets (46percent).

Changing consumer tastes led to a sharp change in theconsumer foods category. Consumption of beef and porkdeclined because of consumer reaction to the sharp priceincreases during the 1970's, and because of health concerns.Consequently, the relative importance of both the beef andveal and the pork categories fell by 10 percent; beef and vealtogether now account for under 2.2 percent of the weightof the total finished goods index, while pork accounts forabout 1.5 percent.

Many Americans substituted poultry and fish for othermeats in their diets. The increased consumption of poultrywas partly due to its relatively low price compared to othermeats. The poultry industry gained a greater share of themarket by reducing costs through improved efficiency inproduction. Poultry presently accounts for 1.2 percent of thefinished goods index, an increase of over 61 percent com-pared to the 1972 weight structure. The relative importanceof fish showed only a small increase.

Weights for confectionery end products rose 27 percent,compared to the 1972 structure, partly as a result of increasedcosts for ingredients, particularly sugar, while soft drinksadvanced 11 percent. Among other foods, the proportionaccounted for by salad dressings rose significantly. Increasesalso occurred for spices, frozen packaged sandwiches, andsnack foods, while bakery products were 17 percent belowthe 1972 weight.

Because prices for finished energy goods more than quad-rupled from 1972 to 1982, declines were experienced in percapita consumption. The net effect was little change inenergy's relative share of consumer expenditures. The grow-ing prevalence of new cars with greater fuel efficiency,coupled with the implementation of a nationwide 55-mile-per-hour speed limit, reduced demand for motor vehicle fuelscompared with the long-term growth trend. Gasoline stillaccounts for about 3.6 percent of the weight of the finishedgoods index, the same as under the previous weight struc-ture. A substantial increase in home insulation efforts, andsome switching to alternatives sources of energy, likewisereduced demand for home heating oil, which now represents0.9 percent of the finished goods index, about the same asbefore. However, the relative importance of natural gas roseto 3.3 percent of the finished goods category, compared to2.1 percent previously; this partly reflected the inclusion ofinterplant transfers in the PPI universe.

Among other consumer nondurables, athletic footwear rose33 percent in importance, as more costly shoes were mar-keted to both fashion- and fitness-conscious consumers; al-coholic beverages increased 29 percent. Increases forplastic dinnerware and tableware reflected the growingimportance and acceptance of these alternatives to chinawareand metal dining utensils. Women's and children's apparelalso gained in importance, while men's apparel declined.Cosmetics and soaps both moved up about 9 percent inimportance.

12

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis

April 1987