Printer Friendly

Industrial production, capacity, and capacity utilization since 1987.

The Federal Reserve recently revised its index of industrial production and the related measures of capacity and utilization. Compared with the previous index, the revised index shows that manufacturing grew more slowly from 1987 until the onset of recession late in 1990 and then recovered more rapidly, albeit unevenly. According to the revised numbers, from 1987 to 1992 the annual rate of growth in industrial production averaged 1.3 percent, 0.4 percentage point less than was formerly shown. From the first quarter of 1992 through the first quarter of 1993, revised industrial output rose at an annual rate of 4.4 percent to a new high. Preliminary estimates show that industrial production changed little in March and April 1993.

The revised production and capacity indexes still cover manufacturing, mining, and electric and gas utilities and are still expressed as percentages of output in 1987. They have, however, incorporated new monthly and annual data-particularly output estimates derived from the Annual Survey of Manufactures of the U.S. Bureau of the Census and capacity utilization results from the Survey of Plant Capacity, also of the Census Bureau. The incorporation of the new data has progressively lowered the level of the revised industrial production index relative to that of the previous one until 1990. Whereas the previous index showed output peaking in September 1990 at 110.6 percent of the level of output in 1987, the revised index shows output reaching a high of 107.1 in April 1989 and then retreating a little until October 1989 (chart 1, top panel). The subsequent rebound recouped nearly all of the loss by the following summer. Then, as did the previous index, the revised index shows a cyclical contraction ensuing. From September 1990 to the following March, production dropped 4 percent to a low of 102.5, compared with the previous index of 105 percent. By February 1993, industrial output had reached 109.9 percent of its 1987 average--a new high, but 2.1 percentage points below the unrevised level.

The Federal Reserve's index of industrial capacity was also revised downward. According to the revision, the annual growth in capacity from 1987 to 1992 averaged only 1.7 percent, down from the previous estimate of 2.3 percent. The rate of capacity growth has trended lower for many years; it tends to be especially slow during and immediately after recessionary periods, when plant closings increase and capital spending drops.

As revised, capacity utilization, the ratio of output to capacity, was about the same for most of 1987-91 as it had been before the revision. It peaked at 84.8 percent in March 1989 and fell to a cyclical low of 78.3 percent two years later. Neither the high nor the low reached the extreme values of the two preceding business cycles. During 1992, utilization rose with production and reached 81.4 percent in the first quarter of 1993, 0.5 percentage point below its 1967-92 average. However, the more rapid recovery of revised production from the trough and the slower growth of revised capacity led to an upward revision of utilization in 1992 and early 1993. For total industrial production, capacity, and capacity utilization for the period of the revision, see table 1. [TABULAR DATA 1 OMITTED]


Pronounced cyclical contractions and recoveries have been most evident in the output of construction supplies and consumer durable goods, particularly automotive products (chart 2). From its high in early 1989 to the trough in early 1991, output of consumer durable goods fell about 20 percent. After an uneven recovery, it regained its previous high in early 1993. Because of the peaking of building activity in 1986, production of construction supplies began to decline before the general cyclical peak, (table 2). Today, despite the lowest interest rates in nearly a quarter century, the recovery in output of construction supplies continues to be constrained--at least in part--by the high vacancy rates of rental housing and commercial buildings. In contrast to these strong cyclical patterns, the output of nondurable consumer goods, business supplies, and energy materials has fluctuated relatively little in recent years. [TABULAR DATA 2 OMITTED]

The production of total equipment grew about 3 percent a year on average from early 1987 to early 1993, even though output of business vehicles and heavy machinery fell sharply in the recession and production of defense equipment has been cut considerably since 1990. The recently ended boom in commercial aircraft and the strong gains in output of computers, communications equipment, and medical instruments boosted the business equipment group in the 1987-92 period. From early 1992 through early 1993, output of business equipment soared at an annual rate of more than 12 percent (8 percent, if office and computing equipment is excluded).

The production of industrial materials used to manufacture durable and nondurable goods, which in the past was more cyclical than that of other industrial products, has, in recent years, roughly paralleled the movements of final products (consumer goods and equipment) production.(1) This parallel movement may reflect the tighter management of stocks of materials that reduced overbuilding and the need for cutting stocks. Swings in net imports of some materials may have counterbalanced some of the swings in domestic demand.


Although industries within manufacturing, mining, and utilities face similar cyclical forces, they are also influenced by other factors, such as technology and resource availability, that may be specific to an industry. As a result, short movements and trends in output, capacity, and utilization differ widely among industries.


For analytical purposes, the large and diverse manufacturing sector--85 percent of total industrial production--is divided into two broad categories: primary-processing industries, which produce mainly materials and construction supplies, and advanced-processing industries, which produce mainly finished consumer or capital goods.

Primary-Processing Industries

Among these industries, the output of lumber, stone, clay and glass, and primary and fabricated metals industries declined relatively severely in 1990-91; output in each case remains below earlier highs (table 3). From a high in January 1989, the output index for primary processing dropped 10 percent to the trough. Although recovery has been substantial since then and output of textiles, paper, and industrial chemicals reached new highs in 1992, growth in output in primary processing industries averaged only 3/4 percent per year from 1987 to 1992. By early 1993, output had approximately recovered its peak (chart 3). [TABULAR DATA 3 OMITTED]

So far in the nineties, capacity growth has slowed to an annual rate of only about 1 percent; the rate of capacity utilization in primary processing, which had peaked at 88 percent in early 1989, recovered from its recessionary dive to a moderate rate of 84 percent in early 1993.

Capacity in some of these basic industries has remained flat. The annual capability for raw steel has stayed near 114 million tons, as older mills have closed and new minimills have been built. Aluminum mills have wrung some additional output of ingots from old potlines, but no new potlines have been constructed. Cement clinker capacity has contracted, and output and capacity in logging camps and lumber mills have been constrained by environmental regulation that reduced logging on millions of acres of federal timberland in the Pacific Northwest (chart 4).

Advanced-Processing Industries

These industries, which account for 70 percent of manufacturing, have grown faster than the rest of industrial production. In early 1992, they regained their peak level; and in the first quarter of 1993, output was 12 percent higher than it was in 1987--up more than 2 percent per year. Even so, utilization remains moderate, as capacity has expanded with output (chart 5). The performance of individual industries in this sector has been quite disparate in terms of cyclical volatility and growth.

Motor vehicles. Led by gains in the demand for trucks, the motor vehicle industry began to recover in 1992 from three years of declining output and sales. In 1991, U.S. output of vehicles, which from 1985 to 1988 had averaged about 11 million units, fell below 9 million units, the lowest level since 1982. Although large, this contraction was not as great as the contractions between 1973 and 1975 and between 1978 and 1982, partly because this cycle had neither a curtailment in gasoline supplies nor so extreme a peak in output. The drop in production of motor vehicles and related parts and materials from its high in January 1989 to the trough accounted for about a third of the decline in manufacturing.

Widely publicized plant closings by Chrysler, Ford, and General Motors accompanied the downtrend in the production of automobiles. The Big Three's auto assembly capacity in the United States fell from more than 9 1/2 million units in 1986-87 to less than 6 1/2 million units in the 1993 model year. In contrast, reflecting the rising share of trucks in overall vehicle sales, U.S. capacity to assemble light trucks increased about a million units to 5 1/4 million units over the same period. Domestic producers shifted existing plants from autos to light trucks, built new truck plants, and increased hours of plant operation. Japanese auto producers continued to increase assembly capacity in the United States; "transplant" assembly capacity (mainly for autos) increased from about 1 million units in 1986-87 to more than 2 million units currently. Altogether, U.S. assembly capacity for motor vehicles has declined about a million units since 1986-87 to about 13 1/2 million units. With the increases in assembly capacity in Canada and Mexico, however, total North American assembly capacity has increased and has been more than ample to meet demand.

The revised output and capacity indexes, shown in the middle panel of chart 6, are based largely on the unit counts mentioned above (left panel); the indexes also incorporate an upward adjustment of about 3 percent per year resulting from gains in the quality of vehicles--antilock brakes, airbags, more efficient engines, more amenities, and so on. The utilization rates for autos and trucks parallel the physical output data and indicate that the rate of capacity utilization, which had dropped from more than 80 percent in the mid-1980s to less than 60 percent at the trough, recovered to about 80 percent early in 1993. The utilization of light truck assembly capacity was much higher than that for automobiles; the capacity utilization rate for medium and heavy trucks also recovered strongly with output.

Aerospace and miscellaneous transportation equipment. Output of transportation equipment aside from motor vehicles dropped 16 percent from the end of 1990 to early 1993, to its lowest level in seven years. Further curtailments in production seem inevitable. Output of aircraft and parts, boats and naval ships, and guided missiles is declining. (The only real strength is in heavyweight motorcycles and mountain bicycles, along with casino boats, ferries, and other smaller craft.) Sharp cutbacks in defense outlays have been the major drag; but the production of commercial transport aircraft, which doubled from 1987 to 1991, is also now falling.

The boom in output of civil transport aircraft peaked in late 1991, although orders had already begun to decline. The boom was fueled by growing demand for air travel. Airlines were hit, however, by decreased air travel in 1991 in response to the Persian Gulf War and recession, by excess capacity, and by substantial price competition in 1992. New orders fell from more than 1,000 units in 1989 to an annual average of 243 units in 1991-92, and backlogs dropped as many airlines cancelled orders or delayed delivery past 1996 in response to record financial losses in 1991 and 1992. Deliveries peaked at about 160 planes a quarter in the first half of 1992 but dropped to 110 in the first quarter of 1993. Airframe makers announced further production cuts for 1993 and 1994; much of the current backlog is not scheduled to be delivered until the last half of the decade. Shipments of complete military aircraft have fallen from about 1,200 units in 1987 to about 600 units in 1992 and 1993.

With the sharp cuts in production, employment and capacity utilization in the aerospace and miscellaneous transportation industries have dropped substantially. The utilization rate fell from 88 percent at its peak in the second quarter of 1989 to 70 percent in early 1993.

Machinery and instruments. The diverse industries that produce machinery, instruments, and parts include fast growers like computers, semiconductors, communications equipment, and medical instruments along with others that produced less at their recent peaks than they had in the early 1980s.

The output of industrial and commercial machinery and computer equipment grew 4.5 percent per year from 1987 to 1992;(2) but the pace was less than half that if office and computing machines are excluded. Because of the phenomenal increase in computers' performance per dollar, the index for real production of computer and office equipment has doubled since 1987. Much of the output gain was registered during 1992 when the U.S. computer industry emerged from a prolonged downturn. In 1991, shipments in current dollars were about 8 percent below their peak in 1988 and only a few percentage points higher than they were in 1987; employment in the computer and office equipment industry dropped 100,000 from its peak in 1984 to 416,600 in 1991.

Excluding computers, output of nonelectrical machinery, which dropped about a tenth from peak to trough, recovered strongly in 1992 but remained below its 1989 high. Makers of engines and of farm and construction machinery produced substantially less in 1992 than in 1980. In 1992, a rebound in residential construction, increased public construction, and the end of a strike at a major producer boosted sales of construction machinery. Nevertheless, because of weakness in commercial construction of office and apartment buildings and the low level of industrial plant expansion, the level of output of construction and mining equipment remains near the weak 1987 level, far below earlier peaks. Although recovering somewhat late in 1992, output of farm equipment also remains below earlier peaks. Fewer farmers, declines in net farm income and the real purchasing power of farm equity, productivity gains, and conservation tillage (which requires fewer trips and reduces wear on the tractor) have reduced the sales of farm equipment.

The production of electrical machinery, which had flattened in 1989-90, advanced sharply from the trough to new highs. Bolstered by gains in semiconductors and communications equipment, output increased at a 3 3/4-percent annual rate from 1987 to 1992. With a double-digit gain in 1992, the production index for semiconductors and related components has risen more than 50 percent since 1987, whereas prices of semiconductors have drifted down. This strong growth is related to the increasing use of semiconductors and related devices in computers, communications equipment, and various recovering industries such as motor vehicles.

The output of communications equipment, which eased a bit in the recession, rose sharply in 1992 to a level more than 25 percent higher than that in 1987. One source of the strength is the robust growth of the cellular radiotelephone system and of fiber optics. During 1992, cellular service became available in each U.S. market block, and the number of new cellular subscribers increased more than 200,000 per month to about 10 million by the end of the year. Net exports of radio and television communications equipment have risen substantially in recent years.

Other components of electrical machinery exhibited less robust growth. The output of appliances and television sets and of electric motors and generators and related parts dropped sharply in the recession and exhibited moderate growth through 1992. In the first quarter of 1993, output of appliances and television sets rose to 121 percent of 1987 output. However, the output of lighting and wiring products, storage batteries, and transformers and switchgear was lower at the end of 1992 than it was in 1987.

In contrast to the cyclical fluctuations of motor vehicles, some machinery, and miscellaneous manufactures, the output of instruments has been relatively flat since 1988. The medical instrument business has been the only strong sector; the output of guidance and navigation equipment has fallen substantially with the decline in defense spending. The rate of capacity utilization for the overall instruments industry is low.

As in the case of instruments, the industrial production index for manufactured nondurable goods has in recent years exhibited a muted cycle and a slow rate of growth. Among advanced processors, the output of foods grew only about 1 1/4 percent per year from 1987 through 1992--about half as fast as in the preceding five years. Changing tastes and lifestyles and reduced family size have shifted production from beef, liquor, whole milk, and large cakes and pies to poultry, beer, skim milk, frozen yogurt, and frozen convenience foods. The production of tobacco products has been trendless for a decade. In contrast, the output of finished chemicals grew more rapidly; this growth was led by drugs and medicines, which advanced a third from 1987 to early 1993.

The industrial production indexes for printing and publishing, paints, apparel, and leather goods, however, were lower in early 1993 than in 1987. A continued uptrend in imports of apparel and leather goods contributed to that decline. The output of printed products was reduced in the early 1990s partly because of a decline in advertising revenues during the recession, which also restrained demand for business publications and forms. A drop-off in corporate takeover activity led to less financial printing. Declines in tax revenues cut school and library budgets for educational materials. Newspaper industry receipts in 1992 declined for the fifth year as newspapers' share of total media advertising and the circulation of daily newspapers shrank further.


From 1987 to mid-1991, mining production, as a whole, changed relatively little, but afterward it declined a few percentage points. The easing was due partly to weak demand for coal, stone, and earth minerals but mainly to the renewed decline in crude oil production and the low rate of oil and gas well drilling that developed as the price of oil retreated after the defeat of Iraq. The estimated index of mining capacity declined as well. Capacity utilization edged down in 1992 and stayed within the 85-90 percent range of recent years (chart 7).

Mining of metal ores increased sharply from the depressed levels of the mid-1980s. New technology and new mines contributed to the doubling of production of gold and zinc ores since 1987; upgraded domestic facilities and disruptions in foreign supplies helped raise output of copper ore in the United States by 40 percent.

Production of energy by mines, refineries, and utilities, representing about 15 percent of industrial production, has been sluggish for the past twenty years. A key factor has been the long-term downtrend in the production of crude oil and natural gas, which has dropped more than 20 percent since 1973. This decline has offset much of the gain in coal mining and nuclear generation of electricity.

After a slight rise in 1991 related to the spurt in oil prices during the Persian Gulf crisis, production of domestic crude oil resumed its decline. From early 1991 to early 1993, it fell 500,000 barrels per day to 7 million barrels per day--the lowest level since 1960. Since 1988, when Alaskan output peaked, the daily rate of production has dropped a million barrels. For more than twenty years, output in the lower forty-eight states has been trending down--dropping from 9.4 million barrels per day in 1970 to 5.4 million in 1992. With the aging of oil fields, oil well productivity has fallen to about 12 barrels per day per well from about 13.5 barrels in 1987-88 and more than 18 barrels in the early 1970s.

Moderate prices of crude oil and environmental restrictions in some areas have depressed drilling activity in the United States. Drilling for oil and gas in terms of either footage or well completions has fallen to less than a third of its peak level in the early 1980s. The number of active or available rotary drilling rigs has also been shrinking.

In contrast to that of crude oil, the output of natural gas, which also declined through 1986, has recovered; and, despite some weather-related weakness since the recession, coal has trended up. The gains in coal mining have been primarily west of the Mississippi.

Refinery production, which had recovered from its lows in the early 1980s, has been basically flat since 1988. Total refinery production of motor gasoline had fallen to 6.3 million barrels in 1982-83 because of the high price of oil and recovered to 6.8 million barrels in 1986-87, when the price of oil fell. From 1988 to 1992, it stayed at 7 million barrels per day, about the same as in the late 1970s. Distillate fuel oil has equalled nearly 3 million barrels per day in the past few years, below the rate in the late 1970s but up from that in the mid-1980s.

In 1992, refinery capacity shrank. More stringent requirements mandated by the Clean Air Act Amendments of 1990 for summer gasoline vapor pressure and oxygenated gasoline, as well as the upcoming reduction in allowable sulfur content of diesel fuels, necessitated the upgrading of existing refineries. Several idle marginal refineries permanently closed. The Department of Energy reports that, at the start of 1993, operable crude distillation capacity dropped to 15.1 million barrels per day, down about 500,000 barrels per day from its average level during 1985-92. Input to crude oil distillation units rose so that utilization increased from 86 percent in 1991 to 88 percent in 1992. This rate is moderate for refining; it is well above the low rates of the early 1980s, when the high price of oil cut demand, but below the rates in the decade before the Arab oil embargo.


From 1986 to 1989, production by electric and gas utilities increased 13 percent; it was spurred by relatively low energy prices and a temporary surplus of deliverable natural gas that caused a rebound in consumption, particularly industrial consumption, of natural gas from its 1986 low. The growth in sales by electric and gas utilities slowed in 1990-92 because of episodes of mild weather and economic weakness. Cool weather caused the generation of electricity to decline in the summer of 1992, but stronger industrial demand for power and a return of more normal weather boosted utility sales in late 1992 and early 1993 (chart 8).

The growth of capacity at utilities in the past few years has kept up with the slow growth of output so that the rate of capacity utilization has changed little. However, the utilization rate is projected to rise. According to the North American Electric Reliability Council, the nation's utilities intend to increase generating capacity at an annual rate of about 1 1/3 percent between 1992 and 2001, compared with a projected annual growth in summer peak demand of 1 4/5 percent. The more rapid rise in demand is expected to reduce the margin of excess capacity a few percentage points.


The downward revision in industrial production was centered in manufacturing, especially in nondurable goods industries other than textiles, paper, and leather goods. The growth of nondurables from 1987 through 1992 was cut in half, to 1 percent per year (table 3). Especially large downward revisions were made to printing and publishing and to paving and roofing materials (part of petroleum products); output in these industries as well as in apparel and leather products was lower in 1992 than it had been in 1987. The output indexes show foods, tobacco products, apparel, and rubber and plastics products growing more slowly than they were previously shown to grow.

In durable manufacturing, several downward revisions, especially to fabricated metal products, instruments, and miscellaneous manufactures, were largely offset by upward revisions to electrical machinery and transportation equipment. For durable manufacturing, the greater rebound from the trough shown by the revised index offset the lower levels in 1989 to 1991, so that by the end of 1992 the revised index exceeded the earlier figures.

For mining, particularly for coal, oil, and gas, the overall 1987-92 revision in output was small. For most series, monthly physical output figures, supplied by the Bureau of Mines or the Department of Energy, were adjusted to more complete annual data from the same sources. This process typically affected specific year-to-year movements rather than the overall 1987-92 rate of growth. In the case of stone and earth minerals, however, monthly output was estimated largely on the basis of kilowatt hours purchased by that industry. These estimates were revised substantially downward to match the annual physical product figures provided by the Bureau of Mines.

The output index for utilities was revised upward because of a correction to the measurement of gas transmission (the output of electric utilities was revised only slightly). Previously the sales of gas by major pipeline companies were used to measure monthly transmissions. However, most interstate pipeline companies have become open-access transporters and, since 1987, have increasingly transported gas owned by others. As a result, sales of gas owned by pipelines fell far below transmissions. In this revision, pipeline sales were replaced by cubic feet of natural gas delivered to consumers. Whereas before the revision a slight decline was shown, the new series shows that gas transmission has not been falling; from 1987 to 1992, output of gas utilities including transmission grew at an annual rate of 2.5 percent.


The growth of industrial capacity from the end of 1986 to the end of 1992 averaged 1.7 percent per year, 0.6 percentage point lower than the rate shown before the revisions (table 4). In manufacturing, large downward revisions to output and capacity indexes are evident in printing and publishing, instruments, and miscellaneous manufactures. As explained below, the downward revisions in capacity growth reflect not only new data on capacity and utilization but also the downward revisions in the industrial production indexes. [TABULAR DATA 4 OMITTED]

The revised capacity indexes are derived from (1) the survey by the U.S. Bureau of the Census of utilization rates in manufacturing establishments for the fourth quarter of years through 1990, (2) output and capacity data in physical units from various sources through 1992, and (3) estimates of capital stock derived through the perpetual inventory method from investment spending by industry.(3) Investment spending comes from the Annual Survey of Manufactures (ASM) through 1991 and from the Census Bureau's quarterly survey of plant and equipment expenditures after 1991.

The Federal Reserve indexes of sustainable annual capacity are designed specifically to accompany the indexes of production and are also expressed as percentages of production in 1987. A basic step in estimating annual levels of the capacity index is dividing an industrial production (IP) index by a utilization rate provided by an outside survey ([U.sub.s]), that is, implied capacity = IP/[U.sub.s] Thus, if in 1987, IP = 100, and [U.sub.s]= 80 percent, then implied capacity = 100/0.8 = 125. For the most part, utilization rates from surveys are revised only slightly, if at all. Thus, the Federal Reserve capacity indexes are typically revised with industrial production indexes, and the Federal Reserve rates of utilization change relatively little on average in a revision (table 5). The sharper increase in 1992 in the revised industrial production indexes, however, has raised utilization above its unrevised level in the fourth quarter of 1992. [TABULAR DATA 5 OMITTED]

This revision was conducted at the end of a very sluggish period. Given the lack of direct capacity measures, the capacity indexes are often inferred indirectly. Making such inferences when output growth has been particularly slow and utilization rates are relatively low has the danger of yielding underestimates of capacity and its growth. Yet, an alternative measure of capacity utilization reported by the National Association of Purchasing Management, Inc., currently shows higher rates of utilization in manufacturing and, implicitly, a lower index of capacity.


In some respects, the revised production indexes look much the same as before. Most noticeably, the 1987 weight and comparison base has been maintained, and the structure of the major market and industry groups has remained basically the same. However, the updating of monthly data and the introduction of comprehensive annual data from the ASM have significantly affected the level and the movement of the indexes.

Many smaller changes were also made in the revision. The structure was modified to conform to the 1987 Standard Industrial Classification (SIC). Some sources of data were lost, and so some series based on those sources were dropped or combined with other series. However, the introduction of new series increased the number of individual series to 255. Monthly indicators for other series also changed. For example, some series that had been based on an industry's monthly purchases of electricity (kilowatt hours) became based on monthly production-worker hours. Seasonal factors were recalculated.(4)

Typically, either monthly output or monthly input data are used to estimate individual production indexes. Monthly output or shipments in physical units from many private or government sources are the basis for 123 series that represent 36 percent of the total index. Input data constitute 74 production-worker-hour series (34 percent of the total index) and 40 kilowatt-hour series (27 percent). Each of these types of monthly data has been updated or revised.

Quarterly production data in physical units have been introduced to maintain the representation of physical measures even though the Census Bureau and other sources have discontinued the monthly publication of some of these data. The quarterly production data are the basis for sixteen series that represent 2 percent of the total industrial production index. A cubic spline generally is used to interpolate monthly values from the quarterly figures.

Revised Federal Reserve indexes of electric power use by industry from 1986 were published in February 1993 and are incorporated in the new industrial production indexes for monthly components based on the use of electricity. The Federal Reserve Banks conduct a voluntary monthly survey of electric utilities, which report sales to industries classified by SIC codes; the respondents also include some industrial cogenerators. The revision reflects conversion of the power series since January 1987 to the 1987 SIC and the incorporation of more comprehensive annual and monthly source data, where available. New power series include indexes for converted paper products (SIC 267) and plastics products (SIC 308) and exclude indexes for SIC 264, 266, and 307, which no longer exist in the 1987 SIC.

The U.S. Bureau of Labor Statistics (BLS) revises its monthly survey data to new levels based on the number of employees covered by unemployment insurance in March. The last available benchmark was for March 1991. In the estimation of the Federal Reserve production indexes, the monthly worker hours (whether or not benchmarked by BLS) are further adjusted to annual production levels derived largely from the ASM.

Estimates of real output derived from the ASM for 1988 through 1991 determine most of the revised annual indexes of production in manufacturing; the Census of Manufactures, a more complete count of the universe, provides indexes for 1982 and 1987 and will do so for 1992. Historically, the ASM figures tend to show less growth than do those from the census. Also, the introduction of a new panel of respondents in 1989 apparently caused an understatement of growth in that year. Consequently, the estimates of output derived from the ASM were adjusted upward by 1/2 percent in 1988 and about 11/2 percent in the years from 1989 to 1991. As shown by table 6, the Federal Reserve's index of manufacturing output has been revised downward to levels close to the adjusted levels derived from the ASM.(5) [TABULAR DATA 6 OMITTED]

Some of the individual production indexes are benchmarked to annual data that are expressed in physical volumes, such as tons of steel or gallons of gasoline, rather than to deflated output figures derived from the ASM. This alternative form of data accounts for some slight difference between the Federal Reserve and ASM-based indexes of output.

Cyclical movements in productivity are used in estimating those monthly indexes based on production-worker hours or kilowatt hours of electricity used by industrial plants. Through 1991, the movements in output per hour can be inferred by adjusting the monthly hours data to more comprehensive annual output measures, such as those developed from the ASM figures. Given the lack of such annual output data since 1991, the cyclical rebound in productivity since then must be estimated from other information such as (1) the movements of output-input ratios in previous cyclical recoveries and (2) the productivity changes shown in series based on monthly output data in physical units. Other indicators of current production, such as manufacturers' shipments and inventories in constant dollars, are used to corroborate the current trends in output-input ratios.

Because the revised industrial production indexes follow the 1987 SIC, a number of changes--including splits or recombinations of series--occurred. One change involved the shift of some military equipment from SIC 36, Electrical Machinery, into SIC 38, Instruments; this shift lowered the growth of this relatively small industry group more noticeably than it raised the growth of the larger electrical machinery group.(6)

The separate series on government-owned and -operated army, air force, and navy ordnance facilities and navy shipyards were eliminated because the ASM and the Census of Manufactures provide no annual data for benchmarking the available employment data.

In calculating new seasonal adjustment factors, the staff applied the X-11ARIMA program of Statistics Canada to data through October 1992. Seasonal factors were forecast twelve months ahead to October 1993. In cases in which cyclical movements appeared to distort seasonal factors, some original values were replaced with maximum likelihood estimates of more typical values before running the X-11ARIMA program. A prior adjustment for Easter, which may fall in March or April, was also made for some physical product or kilowatt-hour series. The Easter adjustment factor was later combined with the factor computed by the X-11 to get the final seasonal factor. For production-worker hours, some prior adjustment may have been required when holidays, specifically Easter and Labor Day, occurred during the pay period containing the survey week for the BLS employment survey. (1.) The total products grouping includes consumer goods, equipment, and construction and business supplies. (2.) Standard Industrial Classification (SIC) 35, formerly called Nonelectrical Machinery. (3.) A methodology of the capacity and utilization measures is found in Richard D. Raddock, "Recent Developments in Industrial Capacity and Utilization," Federal Reserve Bulletin, vol. 76 (June 1990), pp. 411-35, especially pp. 424ff. One of the key sources is the US. Bureau of the Census: "Survey of Plant Capacity," Current Industrial Report MQ-C1, for various years. This survey provides utilization rates, not capacity estimates, for manufacturing establishments at the four-digit SIC level. (4.) A table "Industry structure and series composition of industrial production: classification, value added weight, and description of series" is available with a diskette of historical data of industrial production, capacity, and capacity utilization from Publication Services, Board of Governors of the Federal Reserve System, Washington, DC 20551. A typed structure table may be obtained from the Industrial Output Section, Division of Research and Statistics. (5.) Gross output at the four-digit SIC industry level equals value added plus the cost of materials. The current dollar figures from the ASM are deflated, for the most part, by the BLS's producer price indexes. The main exception is the hedonic price index used to deflate computers. The deflated four-digit industry output figures are indexed on a 1987 base, weighted with proportions based on value added in 1987 reported in the Census of Manufactures, and aggregated. This is essentially the same method of aggregation used to combine the Federal Reserve's indexes of production. (6.) More specifically, search, detection, navigation, and guidance systems and equipment, part of old SIC 3662, were transferred to part of new SIC 3812; hydrological, hydrographic, meteorological, and geophysical equipment, (also part of old SIC 3662) to part of SIC 3829, Measuring and Controlling Devices, not elsewhere classified; and X-ray apparatus and tubes (old SIC 3693) to SIC 3844,5.

Availability of Output, Capacity, and Utilization Estimates

Current estimates of output, capacity, and utilization are published first in the Federal Reserve statistical release G.17(419), "Industrial Production and Capacity Utilization," and then in the statistical tables of the Federal Reserve Bulletin. All data shown in the release will be available on the day of issue through the Department of Commerce's online Economic Bulletin Board (202-377-3870).

Diskettes containing either historical data (through 1985) or revised data (from 1986 to those most recently published in the G.17 statistical release) are available from Publications Services, Board of Governors of the Federal Reserve System, Washington, DC 20551 (202-452-3245). Estimates for total industry and manufacturing, January 1981 to the latest available month in 1993, are shown in tables 5A and 5B of the G.17 statistical release. Hard copy of the revised estimates of series shown in the G.17 release is available upon written request to Industrial Output Section, Mail Stop 82, Division of Research and Statistics, Federal Reserve Board, Washington, DC 20551. A table "Industry structure and series composition of industrial production: classification, value added weight, and description of series" is also available upon written request to the Industrial Output Section. A similar table is available for the capacity indexes.
COPYRIGHT 1993 Board of Governors of the Federal Reserve System
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:Raddock, Richard D.
Publication:Federal Reserve Bulletin
Date:Jun 1, 1993
Previous Article:Reserve requirements: history, current practice, and potential reform.
Next Article:Industrial production and capacity utilization for March 1993.

Related Articles
Industrial production and capacity utilization.
Industrial production and capacity utilization.
Industrial production and capacity utilization.
Industrial production and capacity utilization for April 1993.
Industrial production and capacity utilization for March 1994.
Industrial production and capacity utilization for October 1995.
Industrial production and capacity utilization for April 1996.
Industrial production and capacity utilization: annual revision and 1997 developments.
Industrial Production and Capacity Utilization: 1998 Annual Revision.
Industrial production and capacity utilization: the 2001 annual revision.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters