08/14/2026

The cost of doing business: input shares and trends in input-related price indexes

This article analyzes input shares data from the U.S. Bureau of Labor Statistics (BLS) productivity program to compare how reliant different sectors were on capital, labor, energy, materials, and services inputs from 2019 through 2023. The article also highlights trends in the Producer Price Index (PPI) program’s intermediate demand indexes, alongside other BLS input-related indexes, from December 2018 to December 2023. December 2018 is when BLS began publishing its experimental satellite inputs to industry indexes, and the following 5-year period can be used to analyze the impacts of the COVID-19 pandemic and other events on inflation. The input shares and price data can be used together to better understand how real-world events impact final demand prices. Data users may replicate the analytical framework introduced in this article with more recent and industry-specific data for timelier insights specific to their needs.

Generally, inflation is referred to as the rate of increase in prices over a given period and can reflect changes in prices faced by producers and consumers. After decades of moderate consumer inflation as measured by the BLS Consumer Price Index (CPI), recent years have brought more volatility in prices. But inflation is not a single number—it is the impact of price changes throughout the production process across many products’ markets.

Although prices for goods and services purchased by households, as measured by the CPI, are an important and highly visible part of the national inflation picture, changes in consumer prices are often driven at least in part by changes in businesses’ input costs. Less visibly, many goods and services are sold primarily to other businesses as inputs to production, also known as intermediate demand, and commonly referred to as business-to-business transactions. Price movements and trends based on these transactions are included in the Producer Price Index (PPI).

For transactions included in both the producer and consumer price indexes, other factors, such as industry concentration (competition), profit margins, and the necessity or substitutability of a product (elasticity of demand), determine the extent to which changes in production costs are passed on to buyers. Given these other variable factors, BLS data can shed a light on how changes in production costs are transmitted from earlier stages of production to final demand.

In this article, we focus on the producer. First, we compare large sectors of the economy in terms of their reliance on the five input categories—capital, labor, energy, materials, and services (KLEMS).¹ Then, we consider intermediate demand and other input-related indexes that roughly map to these input categories and examine their movements for the total economy and for the manufacturing sector. These higher-level data can inform what underlies price changes for broad categories of final demand products, with implications for policymakers and businesses that consider overall macroeconomic trends. Finally, we consider similar input share and price index data for more detailed subsectors, which may be of greater interest to businesses and industry associations.

Comparing input shares across sectors of the economy

The BLS productivity program’s input shares data show how reliant a sector of the economy is on the five categories of inputs to production: capital, labor, energy, materials, and services (KLEMS). Energy, materials, and services inputs are sometimes grouped together as “intermediate inputs,” which unlike capital and labor are used completely in the production process—an important distinction as we consider different input-related price and labor cost indexes later in the article. 

Using the North American Industry Classification System (NAICS), chart 1 breaks out the input shares of different sectors of the United States economy. These data are published as annual values, and though fairly stable over time, can reflect changes in business operations. To account for year-to-year movements and better align with the forthcoming price index analysis, 5-year averages from 2019 through 2023 are used.

As shown in chart 1, goods-producing sectors tend to rely more on materials inputs, whereas service-providing sectors tend to rely more on labor inputs. For example, materials accounted for an average of 31.2 percent of the manufacturing sector’s inputs over the 5-year period, compared with a 10.8-percent share in the aggregate services sector (NAICS 54-81). Labor accounted for only 27.2 percent of the manufacturing sector’s inputs, compared with 55.1 percent for the services sector.

However, labor intensity is not uniform throughout the service-providing sectors (which includes, but is not limited to, NAICS 54-81). Some service-providing sectors had labor shares closer to the manufacturing sector than the aggregate services sector (NAICS 54-81), such as information (25.8 percent) and financial activities (25.3 percent). Drilling down further into the aggregate services sector by its component sectors, we see similar variation, with the labor inputs’ share ranging from 31.3 percent for educational services to 60.1 percent for professional, scientific, and technical services. (See chart 2.)

While the KLEMS categories are generally straightforward in what they represent, there are some technical nuances. First, costs tied to workers from employment services firms (such as temporary help agencies) are considered a service input rather than a labor input. For industries that rely more heavily on these employment services, the data may show a higher services share and lower labor share than expected.

Second, energy is an important but perhaps underrepresented input in charts 1 and 2. The energy input shares only represent the energy directly used in production, which were highest in the utilities and transportation and warehousing sectors (22.9 and 8.4 percent, respectively). However, energy is unique in that it is also indirectly used as an input in the production of virtually every other input. This attribute of energy’s role in production, along with its higher weight in the CPI, helps explain the public attention paid to energy prices despite the relatively low energy input shares for most sectors.

Next, consider the cost associated with the physical space(s) in which business operations occur. Rented real estate is considered a service input, whereas owned structures are considered a capital input. The prevalence of renting versus owning business structures likely varies by industry, firm size, and location (among other factors).

Last, consider the capital input. BLS defines the capital input as the flow of services derived from physical assets (equipment, structures, inventories, and land) and intellectual property used to produce output.² ⁠The derived flow of services is priced based on implicit rental transactions, which account for the deterioration of assets and financial depreciation that occur over time. It is important to distinguish these implicit prices that make up capital’s input share from acquisition prices that occur at the point of sale in market transactions, with the latter being the type measured by BLS price indexes. As such, capital input prices are excluded from the price index analyses in this article.

Trends in high-level input-related price indexes

This section focuses on trends in prices and labor costs that underlie the price movements for final demand outputs in the United States economy. These high-level, or aggregate, indexes include intermediate demand producer price indexes, certain import price indexes (MPI), and the Employment Cost Index (ECI). Note that the components of these high-level aggregate indexes are also published as independent BLS data series.³ For the purposes of this research, trends in these BLS inflation measures are tracked from December 2018 to December 2023.

Chart 3 shows some high-level indexes that represent inputs to production. For intermediate demand producer price indexes, one energy-related, two materials-related (representing unprocessed and processed materials), and one service-related index are included. Note that some high-level intermediate demand producer price indexes, such as those related to food and unprocessed energy materials, are excluded here to focus on inputs used more broadly across sectors of the economy.

The included intermediate demand producer price indexes cover much of the total intermediate demand system representing business-to-business transactions, and their descriptive titles link them to their corresponding input share categories. In line with the technical nuances of the input shares data noted in the prior section, the services for intermediate demand index includes prices for various types of nonresidential rent (such as industrial, office, and retail spaces) and employment services (such as temporary help and co-employment staffing services).

Producer price indexes measure changes in prices received by domestic producers. Prices of imported inputs, as costs of production, are also passed through (to varying degrees) to final consumers. The import price index (MPI) related to industrial supplies and materials is included to account for imported inputs. Imported inputs include both “direct imported inputs,” which are purchased directly by domestic producers as their own inputs, and “indirect imported inputs,” which account for domestic input suppliers’ imported inputs. When considering both direct and indirect imported inputs, imported inputs accounted for nearly 30 percent of intermediate inputs (excluding capital and labor) for the manufacturing sector and about 10 percent for the services sector in 2019.⁴

The Employment Cost Index (ECI) for total compensation for private sector workers represents changes in labor costs. Taken together, high-level indexes of intermediate demand (domestic) producer prices, imported materials prices, and labor costs represent changes for four of the five KLEMS input categories, with capital excluded for reasons noted earlier in the article.

The intermediate demand producer price indexes for unprocessed nonfood materials less energy and for processed materials less food and energy, as well as the index for imported industrial supplies and materials excluding fuels, began to trend upward around mid-2020, as businesses navigated supply chain challenges and shifting consumer demand related to the COVID-19 pandemic.⁵ As the impact of the pandemic subsided, these core materials-related price indexes peaked in the first half of 2022 before trending downward. The intermediate demand producer price index for processed energy goods peaked in June 2022, in part reflecting the impact of the Russia-Ukraine war and related sanctions on crude petroleum and natural gas.⁶

While materials and energy prices began to trend down in the second half of 2022, labor costs represented by the ECI and services for intermediate demand (which relies heavily on labor) continued to trend upward. These increases may seem moderate compared with changes in the materials and energy price indexes, but wages are often thought of as “sticky,” meaning they are slow to adjust to changing market conditions and can exhibit downward rigidity.⁷ Indeed, despite sharper increases earlier in the pandemic, falling materials and energy price levels came close to converging with the still increasing services and labor indexes by the end of 2023.

Price transmission analysis of the manufacturing sector

Changes in prices received by producers, particularly earlier in the production process as tracked by intermediate demand producer price indexes, may foreshadow subsequent price changes for businesses and consumers.⁸ Chart 4 shows intermediate demand producer price indexes and other input-related price and cost indexes to examine the manufacturing sector from unprocessed materials inputs to final demand goods, using an analytical technique known as a price transmission analysis.

Note the exclusion of the producer price index for intermediate demand processed energy goods in chart 4. Energy is not a heavily weighted input category for the manufacturing sector (energy had a 2.0-percent average input share from 2019 through 2023, as shown in chart 1), and as shown in chart 3, the energy price index can be volatile. Excluding this index from chart 4 places greater focus on price and cost indexes that represent more heavily weighted input categories in the manufacturing sector (materials, services, and labor).

Early in the COVID-19 pandemic, many materials input prices—which are heavily used in manufacturing—increased. As noted earlier, efforts to contain COVID-19 disrupted supply chains and shifted consumer demand from services to manufactured goods.

As businesses adapted and pandemic-related restrictions lifted, mismatches between the supply and demand of materials inputs began to ease and consumption patterns began to normalize, putting downward pressure on prices for materials inputs. Prices for unprocessed nonfood materials less energy and for imported industrial supplies and materials excluding fuels peaked in April 2022, whereas prices for processed materials and components for manufacturing subsequently peaked in May 2022.

By summer 2022, increases in the producer price index for final demand goods less food and energy had decelerated. This moderation of core final demand goods prices demonstrates the effects of countervailing forces. While materials input prices began to fall, labor costs for manufacturing employees and trade services for manufacturing industries (the highest weighted service component of intermediate demand specific to the manufacturing sector) continued to trend higher. Despite lower labor and services input shares relative to materials in the manufacturing sector (as shown in chart 1), these input costs and prices tend to be stickier (as noted earlier in the article), which may serve to increase businesses’ sensitivity to them in pricing decisions.

Analyzing more detailed subsector input shares and inputs to industry indexes

As with the sectoral level, BLS calculates and publishes input shares data for most subsectors (3-digit NAICS) and for select industry groups (4-digit NAICS). Chart 5 highlights the similarity in input shares for ambulatory healthcare services and food services and drinking places—two otherwise unrelated subsectors.

Despite the similarity in input shares for ambulatory healthcare services and food services and drinking places, their actual inputs are naturally quite different given the different services each subsector provides. Beginning with December 2018 data, BLS began calculating and publishing experimental satellite inputs to industry indexes for most 3-digit NAICS subsectors, enabling data users to easily track changes in a subsector’s input costs.⁹ Notably, these satellite inputs to industry indexes include imported inputs, so additional MPI indexes are not needed at this level of analysis. However, these indexes only cover intermediate inputs (excluding labor and capital), making input shares important supplemental data in understanding the percent of total inputs the satellite indexes (and ECI for labor) cover.

Note that while chart 6 shows the different rates at which the two subsectors’ input costs changed, these indexes cannot be used to compare the change in each subsector’s input costs in dollar terms. The point is not to compare the two subsectors’ input costs in dollar terms, but rather to show they moved differently over the period of analysis, reflecting how real-world events impacted their specific inputs and labor markets (the reasons for which are outside the scope of this article).

Satellite inputs to industry indexes are typically published alongside three subindexes: domestically produced goods, domestically produced services, and imported goods, which can also be mapped to the KLEMS input categories.¹⁰ In addition to providing greater insight into the movements of aggregate satellite inputs to industry indexes, each subindex is published with a corresponding relative importance value, which represents the percent that the subindex accounts for in a subsector’s overall intermediate inputs. For example, based on December 2023 relative importance values, ambulatory healthcare services’ intermediate inputs consisted of 28.296 percent domestically produced goods, 61.193 percent domestically produced services, and 10.511 percent imported goods.

Like the high-level indexes considered earlier, the satellite inputs to industry indexes and their subindexes are weighted aggregations of more detailed BLS price indexes, whose movements reflect the dynamics of their specific product markets. These components and their relative importances are listed in the input index relative importance table, which, in addition to providing even more detailed insight into the movements of the aggregate satellite indexes, can be used to inform decisions related to contract price adjustment clauses.¹¹  

Conclusion

From December 2018 to December 2023, prices were impacted by the COVID-19 pandemic and, to a lesser extent, by the Russia-Ukraine war. While these were larger events that occurred over this period, a multitude of factors are constantly affecting markets and prices.

The Producer Price Index, one of the oldest economic time series compiled by the federal government (first published in 1902 covering data back to 1890), is a trusted source of information on prices received by producers, and its monthly publication makes it a timely resource for data users.¹² This article expands on the PPI program’s well-established uses by exploring how more recently developed producer price indexes for intermediate demand, as well as other BLS input-related indexes (satellite inputs to industry indexes, MPI, and ECI) provide greater insight when considered alongside the BLS productivity program’s input shares data.

Examining the reliance of sectors or subsectors on KLEMS input categories, and monitoring the changes in input prices and labor costs related to those categories, can offer valuable insight into how final demand prices may change in the future. For policymakers and businesses interested in macroeconomic trends, higher-level sectoral data may inform economic forecasts. The more detailed subsector data may help businesses make more informed operational decisions through input cost analyses or by informing contract adjustment clauses.

Examples used in this article are representative of BLS data availability and chosen for illustrative purposes. One of the strengths of the data highlighted here are their breadth; similar, timelier analyses can be done for any sector and most subsectors of the economy, and BLS representatives are available to assist with more information as needed.