Methodology

Methodolgical Overview

The TSAs measure the magnitude of in-house transportation services by estimating the inputs used by each industry for its in-house transportation activities. To do this, the TSAs rearrange the I-O accounts using data on transportation from other sources. The following describes the steps involved. The major data sources are identified in table 10.

Step 1. Estimating Transportation Inputs

Transportation inputs include both intermediate inputs (e.g., motor vehicle gasoline) and value-added inputs (e.g., employee compensation, indirect business tax and nontax liability, and consumption of fixed capital). The value of these inputs for each industry in the U.S. I-O accounts' use table combines that for transportation and nontransportation purposes. The TSAs separate the value from transportation uses from all other uses through the following steps:

• Identifying the Transportation-Related Inputs and the Dollar Value of In-House Transportation Services: The TSAs define a set of inputs unique to or mostly used for transportation by a specific mode. These inputs, derived from the underlying items in the I-O accounts, are called "transportation-related inputs" (TRIs). The TRIs used in the 1997 TSAs are presented in table 11.
  
  TRIs may not be used exclusively to produce transportation services or by a specific mode. For example, motor gasoline is required for truck transportation operations. However, motor gasoline is not a TRI exclusive to truck transportation as it is used by other transportation modes and other industries for both transportation and nontransportation-related purposes. For example, motor gasoline is used to carry cargo by watercraft and to operate nontransportation-related machinery. In these cases, the component used for nontransportation purposes, such as gasoline used for heating or for operating machinery, and the component used for transportation modes other than the one for which the TRI was selected, such as gasoline used for transportation modes other than trucking, is removed. The resulting value is used to estimate the total value of in-house air, rail, truck, and water transportation.
   
• Developing Industry Distribution Weights and Distributing TRIs: To measure each industry's production and use of in-house transportation, the value of in-house transportation for each mode (calculated in the step above) must be distributed across industries. This is done through distribution weights. The weights assign a larger portion of the dollar value of in-house transportation to industries with a larger proportion of the total stock of transportation vehicles (aircraft, railcars, trucks, and water vessels) for a given mode. This assignment is based on the assumption that industries with a larger stock of transportation vehicles produce and consume more in-house transportation. Stock values for the 1997 TSAs were compiled from the sources listed in table 10.
   
• Estimating Other Inputs: Transportation activities require inputs not unique or primary to transportation. For example, office supplies and accounting services are shared across transportation and all other production activities. The TSAs separate transportation and nontransportation use of these inputs by assuming nontransportation industries use these nontransportation related inputs in the same proportion as for-hire industries.¹⁸

Technical documentation containing details on the above steps is available on request: at https://ntl.custhelp.com/app/ask

In the above steps, the TSAs assume the following:

• the TRIs are representative inputs for the production of in-house transportation services,
• the selected distribution weights are reliable predictors of the use of TRIs for in-house transportation, and
• the distribution of nontransportation-related commodity inputs within a for-hire transportation industry is similar to the distribution of the inputs within a nontransportation industry for its transportation-related activities.

Step 2. Deriving the TSAs Make and Use Tables

The second step in the TSAs involves the production of the make and use tables. As noted earlier, the TSAs' make and use tables are I-O make and use tables modified to show estimates of transportation related inputs. First, estimates of transportation inputs for each industry are arranged in a matrix so that the rows and columns correspond to those in the intermediate industry portion of the I-O use table.¹⁹ Second, the estimates in the transportation input matrix are subtracted from the corresponding elements of the I-O use table, resulting in a residual use table that shows the intermediate and value-added inputs used by industries for nontransportation activities. Third, the TSAs use table is derived by combining the residual use table, an in-house transportation column vector for each in-house mode with the row sums from the transportation matrix, an in-house transportation row vector for each in-house mode with column totals from the transportation input matrix, and the final-demand portion of the I-O use table. Finally, the TSAs make table is formed by adding an additional column and an additional row for each in-house mode to the I-O make table. The values for this column and row are derived from row and column totals respectively from the transportation related input matrix.

Changes In Method In the 1997 TSAs and Comparability

The steps in measuring in-house transportation services in the 1997 TSAs follow, in general, those used in the 1992 and 1996 TSAs. However, the 1997 TSAs exclude some of the TRIs used in the 1992 and 1996 TSAs and use a slightly different technique for estimating other inputs used to provide transportation services. Additionally, the 1997 TSAs measure in-house transportation services for four modes (air, rail, truck, and water) and provide separate estimates for each. The 1992 and 1996 TSAs measured in-house transportation services for only two modes (bus and truck operations) and combined the two into a single measure.

The 1997 TSAs also differ significantly from the 1992 and 1996 TSAs because of a change in the industry classification system used in the I-O accounts. Prior to 1997, the I-O accounts were based on the Standard Industrial Classification System (SIC) and since 1997, have been based on the North American Industry Classification System (NAICS). NAICS differs from its predecessor SIC in that it:

• classifies industries according to similarities in their production process rather than similarities in the primary products produced,
• treats auxiliaries as establishments and hence explicitly measures the economic activities of auxiliaries rather than including them in the value added of the industry using the services,
• provides additional detail for the services sector, and
• introduces the "information sector".²⁰

The changes introduced by NAICS have resulted in most industries in the 1997 benchmark I-O accounts not being comparable to those in prior I-O accounts and, hence, in most industries in the 1997 TSAs not being comparable to those in prior TSAs.

To enable comparisons between the 1997 TSAs and 1992 TSAs, this report presents revised numbers for the 1992 TSAs using the 1997 TSAs procedure at the sector level. The 1996 TSAs were not re-estimated using the 1997 TSAs procedure due to a lack of sufficient detail in the inputs.

¹⁸ Adjustments were made after applying the for-hire relationship to ensure that the value of the transportation and nontransportation component of each commodity used by an industry equals that in the I-O use table.

¹⁹ Inputs for for-hire transportation industries in this matrix are all zero because these industries, by assumption, do not have any in-house transportation activities.

²⁰ For information on the treatment of secondary products in the 1997 U.S. I-O accounts, see: Ann M. Lawson et al., "Benchmark Input-Output Accounts of the United States, 1997," Survey of Current Business, 2002, 82(12), p. 2, available at http://www.bea.gov/scb/pdf/2002/12December/1202I-OAccounts2.pdf as of Mar. 15, 2011.