Household-Provided Transportation: An Extension of the Transportation Satellite Accounts

Household-Provided Transportation: An Extension of the Transportation Satellite Accounts

Duanjie Chen
Bingsong Fang
Xiaoli Han

MacroSys Research and Technology

Brian W. Sloboda*
Bureau of Transportation Statistics U.S. Department of Transportation


Currently, transportation statistics provide estimates for household demand for transportation services. Though important, there is a missing link to the household supply of transportation services. Incorporating the household supply with household demand provides a complete analytical model for understanding the effects of household use of transportation services. The household production of transportation services is incorporated into the analytical framework of the Transportation Satellite Accounts, which maintain strong ties to the U.S. Input-Output Accounts and the National Income and Product Accounts. Our results indicate that the contribution of transportation activities to total gross domestic product, including the household sector, is 11.6% compared with 5.0% by the Transportation Satellite Accounts and 3.1% by the U.S. Input-Output Accounts. These results for household-provided transportation reveal the importance of this sector in transportation services.


Transportation statistics often treat households as users rather than producers of transportation services. In reality, the household production of transportation services (HPTS) is ubiquitous in our economy, as people drive automobiles (including rental cars) to commute to work or school, go shopping, or obtain commercial and government services.1 Consequently, households produce transportation services to meet their transportation needs, which include their purchase of for-hire transportation services such as ground passenger transportation (including taxicab) and railroad services.

Without an adequate evaluation of household production, or supply of transportation services, the conventional statistics on household demand for transportation services are not complete. Moreover, HPTS has a strong tie not only to the for-hire transportation industry (in both a complementary and substitutive manner) but also to other transportation-related industries ranging from auto manufacturing to the entertainment sector (mostly in a complementary manner). Therefore, a well-defined HPTS industry and properly estimated economic value of HPTS are critical to a better understanding of the overall level of transportation activities and their interaction with other sectors of the economy.

Many efforts have been made to better understand the true magnitude of overall transportation production in the economy. For example, the Bureau of Economic Analysis (BEA) in the U.S. Department of Commerce publishes personal consumption expenditures (PCEs) on user-operated transportation through the National Income and Product Accounts (NIPA)(USDOC BEA 2002, tables 2.4 and 2.5), and the Bureau of Transportation Statistics of the U.S. Department of Transportation publishes vehicle-miles traveled (vmt) by passenger cars (USDOT BTS 2002, table 1-28). A concise presentation of major measures of transportation's economic importance can be found in Han and Fang (2000). Unfortunately, none of these statistics or measures looks at households from a producer's perspective. As such, they do not provide an adequate measure for HPTS that is consistent and hence comparable to the for-hire transportation industries and other productive components of our economy.

Existing statistics either overlook the labor input to household transportation activities (i.e., time spent driving); make no distinction between household capital expenditures on automobiles such as the purchase of vehicles, and household operation and maintenance expenditures for automobiles such as repairs, oil changes, and other related expenditures; or provide no direct monetary value for comparison across economic sectors (e.g., vmt by passenger cars do not imply the same economic value across different type of motor vehicles). These statistics, nevertheless, provide helpful ideas and basic data for measuring HPTS as an "industry" so as to make it comparable to transportation industries that are accounted for in the NIPA. By measuring HPTS as an industry, we are able to extend the existing Transportation Satellite Accounts (TSAs; see box) to include another source of transportation services—HPTS—which is a step further toward estimating the true magnitude of overall transportation in our economy. In the present analysis, transportation services in the form of driving by the household sector are considered regardless of what is transported (e.g., passengers or freight.) We then estimate the economic value of the for-hire transportation industries that provide transportation services either as intermediate inputs to other business sectors or as a final product used by consumers. We present the methodology for this estimation as well as the results for measuring HPTS as an industry, which is an extension of the existing TSAs.

Transportation Satellite Accounts

In general, satellite accounts are frameworks designed to expand the analytical capacity of the basic economic accounts without overburdening them with details or interfering with their general-purpose orientation. Satellite accounts are meant to supplement rather than to replace existing accounts and organize information in an internally consistent way that suits a particular analytical focus, while maintaining links to the existing national accounts. They typically expand a particular segment of existing accounts with more details and additional dimensions of information, including nonmonetary information. They also may use definitions and classifications that differ from those in the existing accounts. Depending on the analytical focus, the production boundary of the national accounts can be maintained or modified (USDOT BTS 1999). The existing Transportation Satellite Accounts (TSAs), as a supplement to the U.S. Input-Output (I-O) Accounts, are developed to cover both for-hire transportation services (identified as transportation industries within the U.S. I-O Accounts) and in-house transportation services conducted by businesses for their own use (not separately identified as transportation in the I-O Accounts). Therefore, compared with the U.S. I-O Accounts, the existing TSAs provide a more comprehensive measure of transportation services provided by the business sector of our economy. Hence, the household production of transportation services extends this measure to transportation services by the nonbusiness sector of the economy, namely households.


Measuring HPTS as an industry expands the production boundary. To count HPTS as an industry requires an estimation of the labor and capital inputs, which are likely to increase the total value added or gross domestic product (GDP). More specifically, households' driving time that currently has no value in the national accounts needs to be counted as labor input to HPTS.2 Furthermore, household purchases of automobiles, which are identified as PCEs in the national accounts, need to be redefined as fixed capital expenditures to derive the capital input to HPTS. While the capital input to HPTS may be estimated on the basis of national wealth data, which mainly involves a technical rearrangement, estimating labor input requires major conceptual clarifications.

The following clarifications are aimed at answering two questions. First, why is household driving, apart from general household traveling, considered a productive activity? Second, how should this productive activity be valued as a labor input to HPTS?

Using the third-person criterion. Horrigan et al. (1999) define the third-person criterion as an activity deemed productive if it can be delegated to another person and still achieve the desired result. According to this criterion, driving is considered productive and hence may be assigned values; in contrast, riding as a passenger is not productive (i.e., it cannot be delegated to someone else) and hence should not be valued.

Using the input approach to estimate the value of driving time. The two conventional approaches to measuring nonmarket work are the input approach and the output approach. The output approach requires directly assigning a monetary value to an output, based on the price at which the output can be sold in the market. This approach is impractical in our case because there is no obtainable price in the market for HPTS; that is, transportation services produced by households are for their own use, not traded in the market.

The input approach is simpler and requires only an estimate of the amount of time a nonmarket activity takes and the selection of an appropriate wage for the specified activity. In our case, the driving time may be estimated on the basis of statistics of average driving speed and the total vmt by household-operated automobiles. Such statistics are available from several government publications as detailed in the appendix table A1.

Using a "generalist" wage as the wage rate. There are three possible wage rates that can be used to estimate the value of household driving time. They are: 1) the individual's occupational wage rate (e.g., the average wage rate for a lawyer, a secretary, or a construction worker), 2) a specialist wage rate (e.g., a racing car driver's rate), and 3) a generalist wage rate (e.g., a taxi driver's rate).

Out of the three types of wage rates, an individual's occupational wage rate serves as the best measure of the driver's true opportunity cost. However, treating the individual's occupational wage as his true opportunity cost not only imposes a very stringent data requirement but also implies strong assumptions concerning optimality in the individual's time allocation and the flexibility of market demand for any type of working hours. In other words, to use the individual's wage rate, one has to obtain a complete occupational profile of household drivers that is also classified by driving time. Furthermore, this approach assumes that the individual driver would desire to spend any time saved from driving on making additional income at his/her occupational wage rate (and that would be permitted by a horizontal demand curve for each occupation). Obviously, both the data requirements and the stringency of the conceptual assumptions required make this approach impractical.

It should be noted that using a generalist wage rate such as a taxi driver's average hourly earning poses a conceptual problem similar to the case of using an individual's occupational wage rate. That is, it is not likely that time saved from individual driving would be used to earn extra income at a generalist wage rate. However, using a generalist wage rate imposes far less difficulty in terms of data requirements. It also makes our estimate of labor input to HPTS relatively conservative since a taxi driver's average earnings are typically below average earnings of nonfarm production workers.3

Procedures for Estimating HPTS

We took a two-step approach to the estimation process. The first step estimates the intermediate inputs and the second, the value added. All the data sources used for our estimate are summarized in the appendix.

Step 1. Estimating the Intermediate Inputs for HPTS. Inputs for HPTS may be grouped into three categories: capital inputs, labor inputs, and intermediate inputs. Capital inputs represent the depreciation of transportation equipment such as household-owned automobiles. Labor inputs are the value of the driver's time. Finally, the intermediate inputs are the goods and services consumed in the process of providing transportation services, including PCEs on tires, accessories, and other parts; repair, greasing, washing, parking, storage, rental, and leasing; gasoline and oil; bridge, tunnel, ferry, and road tolls; and insurance. In the national Input-Output (I-O) Accounts, such PCE items are considered part of final demand, reflecting the general national income accounting practice that they are purchases for the purpose of consumption rather than production. In order to generate an HPTS sector within the production boundary of the national accounts, we must reclassify these PCE items as intermediate inputs to the HPTS sector.

Step 2. Estimating the Value-Added Inputs for HPTS. Generally, value-added inputs include the value of labor services, indirect taxes, and other value added, mainly value of capital services. The value of labor services is the compensation paid to labor; the indirect taxes include any taxes paid during the process of production, such as import tariffs and license fees; and the value of capital services consists mainly of two parts: capital depreciation and the return to capital.

The value added by labor services, as discussed earlier, is estimated by applying the generalist wage rate (e.g., taxi drivers' average hourly earnings) to driving time. The driving time component of HPTS is estimated as the total household vmt divided by the household vmt per hour, both of which are based on data from the 1995 Nationwide Personal Transportation Survey (see Hu and Young 1999) and other highway statistics (e.g., vmt by type of vehicle).

The indirect tax and nontax liabilities component of HPTS (e.g., motor vehicle license fees) can be obtained from the Personal Income and Outlays data in the NIPA. However, it should be noted that, in the I-O Accounts, indirect taxes such as license fees related to HPTS are not presented explicitly but are contained in the "compensation of employees" by industry. Therefore, to avoid double counting, the indirect taxes component of HPTS is set to zero.

Finally, the return to capital is set to zero to recognize households as a nonprofit sector. This treatment also ignores the time value (i.e., interest income or cost) of money spent by the households on their investment in motor vehicles. Therefore, the value added through capital used for HPTS is mainly the annual depreciation of household-owned automobiles. This depreciation estimate is obtained from BEA, and the exact source is provided in the appendix.


Figure 1 illustrates the magnitude of HPTS in comparison with the business sector of highway transportation (including both for-hire and in-house industries). Figures 2, 3 and 4 compare HPTS with for-hire transportation industries in total output, relative shares within the transportation sector, and annual growth rate. For-hire transportation industries consist of railroads and related services, motor freight and warehousing, water transportation, air transportation, and pipelines and related services. Figure 5 compares the annual growth rate of HPTS with that of government capital stock in highways and streets, educational buildings, hospitals, and other government structures. Figure 6 compares the annual growth rate of HPTS with selected categories of GDP.

Several interesting observations can be drawn from these figures. First, HPTS accounted for a significant portion of total highway transportation services in the economy. As figure 1 shows, in 1999 total vmt by household-owned motor vehicles accounted for over 85% of total vmt by all motor vehicles in operation (including trucks and buses).4 Figure 1 further provides the split between the household and business sectors in net automobile stock (in value terms), fuel consumption, and the motor vehicle license fees paid to governments. This comparison indicates a significant household portion in overall highway transportation activities. Further estimates in figures 2 and 3 show that the value of HPTS was about 1.9 times that of all for-hire transportation industries between 1991 and 2000. Obviously, ignoring HPTS means a significant underestimate of the overall transportation services produced and used in the economy.

Second, household-produced and for-hire transportation services shared a similar growth pattern (figure 4). Figure 4 shows that the growth rates went up and down concurrently. However, the growth rate of HPTS displayed more volatility in comparison to the for-hire transportation sector, especially after 1996. A plausible explanation is that household-produced transportation is more of a discretionary expenditure to the household and more responsive to the ups and downs of the overall economy. On the other hand, for-hire transportation involves more freight movements that are often necessities to the economy and less responsive to the short-term (year-to-year) changes in the growth rate of the economy.

Third, HPTS relies heavily on government investment in transportation infrastructure. To that extent, HPTS is unique among other forms of household production such as food preparation, childcare, and entertainment, because these other forms of household production do not rely on a government supplied infrastructure. It is obvious that HPTS directly consumes the transportation infrastructure as a public good. That is, without government investment in highways and streets, the rapid growth in HPTS would not be possible. Furthermore, HPTS is often related to the household consumption of other major public goods (e.g., schools, hospitals, electric and gas facilities, transit systems, and police and fire stations) through urban sprawl. In other words, urban sprawl fostered the growing need for HPTS. Urban sprawl and hence the need for HPTS might not have grown so rapidly without the government investment in major public goods other than roads (e.g., water and sewer systems). As figure 5 shows, the average annual growth rate in HPTS has closely followed the annual growth rate in government capital stock in highways and streets, education, hospitals, and other nonmilitary structures funded by the government.

Finally, HPTS is one of a few forms of household production that is solidly linked, backward and forward, to many industries. It is obvious that HPTS would not play such a significant role in average people's daily lives if our economy were not equipped with a mature auto manufacturing industry and an efficient system for gasoline distribution. On the other hand, many commercial centers (e.g., clustered consumer centers grouping shopping and recreational facilities) might not have emerged so rapidly without a well-developed system of HPTS. Figure 6 shows that the average annual growth rate of HPTS has been in line with that of manufacturing, retail trade, hotels and other lodging places, and amusement and recreation services. Such illustrations are of course preliminary; a more precise estimate of the economic linkage between HPTS and many industries requires the inclusion of HPTS in the TSAs. As an I-O Account, TSAs facilitate a complete characterization of the interrelationship between one sector and the rest of the economy. To do that for HPTS, however, the TSAs first have to be extended to include HPTS. The next section delves into this.


As mentioned in the introduction, the existing TSAs, as a supplement to the U.S. I-O Accounts, are developed to cover both for-hire transportation services (identified as transportation industries within the U.S. I-O Accounts) and in-house transportation services conducted by businesses for their own use (not separately identified as transportation in the I-O Accounts). Therefore, compared with the U.S. I-O Accounts, the existing TSAs provide a more comprehensive measure of transportation services provided by the business sector of our economy.5 Extending the existing TSAs to incorporate HPTS will allow analysts to measure transportation services provided not only by the business sector but also by the household sector of our economy. In other words, this extension can be seen as a further step toward providing a comprehensive measure of transportation services.

Due to the differences in methodology and data sources between the I-O Accounts and the NIPA,6 certain technical procedures are necessary for the integration of HPTS into the TSAs and for consistency and comparability between the TSAs and the U.S. I-O Accounts. (The details of these procedures are presented in USDOT BTS (Forthcoming)). The following illustrates the structural expansion of the TSAs and the estimated HPTS as additions to each of four TSA tables (i.e., Make, Use, Direct Requirements, and Total Requirements Tables).

Expansion of the TSAs: A Structural Illustration

As mentioned earlier, the inclusion of HPTS as an industry in the TSAs will result in an expansion of the production boundary in the sense that the total value added will be greater than the official GDP. Diagram 1 illustrates such an expansion. The extended coverage of costs and benefits of transportation services enables more accurate analyses of transportation's role in and its contribution to the economy. It also makes it possible to directly link data sources and analysis to the monetary accounting system of U.S. national accounts, which will facilitate transportation analyses in the context of the national economy.

Constructing an HPTS sector within the TSAs follows some basic I-O accounting practices. The overall industry and commodity classification system and the special definitions and conventions in the I-O Accounts7 are used except that HPTS itself forms a new industry and a new commodity category in the TSAs. The general valuation conventions used in constructing HPTS in the TSAs are consistent with those in the I-O Accounts: all transactions including both intermediate inputs and final uses are valued in producers' prices.

Diagrams 2 and 3 illustrate such a relationship more intuitively. Diagram 2, with the shaded industry row and commodity column, depicts the simple fact that the inclusion of HPTS will slightly modify the classification of the existing I-O Accounts. That is, a new industry and a new commodity, both named HPTS, are added to the existing TSAs.

Diagram 3, with further elaborations on how the HPTS industry and commodity are created, shows the expansion of the production boundary resulting from the inclusion of HPTS. More specifically, the user-operated transportation expenditures under PCE in the I-O Accounts and existing TSAs are reclassified from the final demand section to the intermediate inputs section, and GDP is increased by the amount of the total value added of the HPTS industry. For PCE reclassification, user-operated transportation expenditures are treated as either intermediate inputs for the HPTS industry or additions to gross private fixed investment (GPFI), depending on whether they are durable or nondurable goods.

The change in GDP has two sources. First, the value of labor input to HPTS is imputed as a new value-added input. Second, the depreciation of household-owned motor vehicles used for HPTS, which is not counted in the I-O Accounts but contained in the industry and wealth data of the national accounts system, is included as the value of capital services. We discuss these procedures in more detail below.

Addition of HPTS to Each Component
of the TSAs

The structure of the TSAs is consistent with the U.S. I-O Accounts. That is, both accounts are presented in four tables: the Make (production) table, the Use (consumption) table, the Direct Requirements table, and the Industry-by-Commodity Total Requirements table. The addition of the HPTS industry and commodity was made in each of these four tables as one of the three major transportation sectors (i.e., for-hire, in-house, and HPTS). Table 1 provides an overview of major components of the transportation industry and industry output used in the TSAs.

Diagram 2, a TSA Make table, has an additional column for HPTS as a commodity and an additional row for HPTS as an industry. See appendix table A2 where the cell value at the intersection of the column and row of HPTS, under the heading HPTS, equals the total output of HPTS; all other cell entries in the HPTS row and column are zero because no other industries conduct HPTS and the HPTS industry produces nothing but HPTS. The data shown in other parts of the TSA Make table are the same as those provided in the 1992 TSAs, which indicate that the total output of HPTS is a pure addition to the Make table of TSA accounts.

Diagram 3, a TSA Use table, has an additional row for HPTS as a commodity and an additional column for HPTS as an industry (see also appendix table A3, p. 16–18). The entries in the HPTS column are redefined or imputed intermediate and value-added inputs for HPTS. While total intermediate inputs for HPTS are part of the original user-operated transportation expenditures under PCE in the I-O Accounts and TSAs, value-added inputs for HPTS are our imputation as additions to GDP. In this table, the use of the HPTS commodity is shown in the HPTS row. Since the HPTS output is produced and consumed at the same time by households, it is not used by any industries including the HPTS "industry." Therefore, all the cell values in the HPTS row are equal to zero except the one at the intersection of the HPTS row and PCE column in the final uses section.

Besides the additional row and column for HPTS, Diagram 3 also differs from the Use table in the existing TSAs and I-O Accounts in that the part of PCE related to automobile purchases is put under GPFI. Therefore, GPFI shown in Diagram 3 is greater than its counterpart in the I-O Accounts and the existing TSAs. This switch reduces total PCE and increases total GPFI simultaneously and hence does not affect the total output and GDP shown in the I-O Accounts.

As a consequence of these additions to the TSA Make and Use tables, the TSA Direct Requirements and Total Requirements tables also have an additional row and columns for the HPTS commodity and industry. The TSA Direct Requirements table is derived from the TSA Use table by dividing each industry's commodity and value-added inputs by that industry's total output. This table, however, does not include the final use section of the Use table. In the table, each column shows, for the industry named at the head of the column, the input coefficients for the commodities and the value-added components that an industry directly requires to produce a dollar's worth of its output.

The TSA Industry-by-Commodity Total Requirements table shows the total requirements for each industry's output that are directly and indirectly required to deliver a dollar's worth of goods and services to consumers and other final users. Each column shows the commodity delivered to final users, and each row shows the demand for an industry's output in response to a dollar increase in the final demand for a commodity. The coefficients in the table are referred to as industry-by-commodity total requirement coefficients. This table is derived from both the TSA Make and the TSA Use tables.

In summary, the redefined intermediate inputs for HPTS are part of PCE in the existing TSAs and I-O Accounts while the imputed value-added inputs for HPTS are additions. The sum of the intermediate and value-added inputs consumed by the HPTS industry is the total output of the industry, which is consumed by households and hence forms a new category of PCE. As a result, the total output in the expanded TSAs will be greater than the total output of I-O Accounts by the amount of HPTS output plus the in-house transportation industry output, and the GDP in the expanded TSAs will be greater than the GDP of I-O Accounts by the amount of HPTS value added.


The expansion of the TSAs to include HPTS provides a different picture of the magnitude of transportation activities in relation to the economy. Based on the 1992 benchmark I-O Accounts and the 1992 TSAs, three major findings concerning HPTS can be summarized as follows.

True Magnitude of Transportation
in the Economy

The magnitude of transportation in the economy can be measured from two perspectives: as a share of the total output of the economy and as a share in GDP. GDP is the net output of the economy, while total output represents the sum of GDP and all of its intermediate inputs. The share in total output tells more about the size of production, while the share in GDP tells more about its real contribution to the national economy.

The total output of HPTS in 1992 was approximately $710 billion, which is about 1.28 times the total output of business transportation services, including both the for-hire and in-house sectors. Including HPTS in the TSAs would increase the estimate of total output of our economy by about 6.5% and GDP by 7.4%. And the share of transportation (business plus HPTS) in the expanded GDP would be 11.6%.

Use of Major Commodities by HPTS

Counting HPTS as an industry and commodity in the TSAs helps improve the estimate of use of major commodities and services by HPTS. For example, the NIPA counts total PCE on automobiles as a direct expenditure on user-operated transportation. With HPTS treated as a separate industry, the capital input to user-operated transportation can be counted more precisely as the annual replacement portion, or depreciation cost, of the household-owned auto stock that includes current-year purchases.

In the meantime, including all the commodities and services as intermediate inputs to HPTS enables a better understanding of the composition of transportation services provided and consumed by households. For example, within the existing I-O Accounts and the NIPA, one cannot gain any understanding of the services of wholesale and retail trade used by households to meet their needs for user-operated transportation. With the formation of HPTS, services provided by the wholesale and retail trade sector can be shown to account clearly for 26% of total intermediate inputs to HPTS, thus indicating the true significance of the wholesale and retail trade sector in supporting HTPS.

Multiplier Effect of Major Commodities

The multipliers derived from the TSAs including HPTS capture the direct and indirect dependence of HPTS on the rest of the economy. For example, to meet a $1 increase in the final demand for HPTS would require an additional 17.6¢ in total manufacturing output. The next major items for supporting a $1 increment in the final demand for HPTS are services (17.3¢), wholesale and retail trade (10.9¢), finance and insurance (7.7¢), and business transportation services including both for-hire and in-house (3.0¢). These estimates come to light only when HPTS as an industry and commodity is formed within the TSAs. These estimates show a more accurate picture of the household sector in relation to the overall economy.


We are grateful to David Chesser and William Mallett, whose expertise in transportation statistics helped refine our estimate of driving speed, which is critical to estimating the value added by labor service related to HPTS.


Fang, B., X. Han, A.M. Lawson, and S.K.S. Lum. 1998. U.S. Transportation Satellite Accounts for 1992. Survey of Current Business 78(4).

Fang, B., X. Han, S. Okubo, and A.M. Lawson. 2000. U.S. Transportation Satellite Accounts for 1996. Survey of Current Business 80(5).

Han, X. and B. Fang. 2000. Four Measures of Transportation's Economic Importance. Journal of Transportation and Statistics 3(1):15–30.

Horrigan, M., D. Herz, M. Joyce, E. Robison, J. Stewart, and L. Stinson. 1999. A Report on the Feasibility of Conducting a Time-Use Survey, presented at the Workshop on Measurement of and Research on Time Use, May 27–28.

Hu, P.S. and J.R. Young. 1999. Summary of Travel Trends: 1995 Nationwide Personal Transportation Survey. Washington, DC: U.S. Department of Transportation, Federal Highway Administration.

U.S. Department of Commerce (USDOC), Bureau of Economic Analysis (BEA). 1990. Personal Consumption Expenditures, Methodology Paper Series MP-6. Washington, DC: U.S. Government Printing Office. June.

____. National Income and Product Accounts Tables. 2002. Available at, as of Mar. 27, 2003.

U.S. Department of Transportation (USDOT), Bureau of Transportation Statistics (BTS). 1999. Transportation Satellite Accounts: A New Way of Measuring Transportation Services in America, BTS99-R-01. Washington, DC. February.

____. 2002. National Transportation Statistics 2002 . Washington, DC. Also available at

____. Forthcoming. Household Production of Transportation Services Through Automobiles: An Expansion of the Transportation Satellite Accounts . Washington, DC.

Yuskavage, R.E. 2000. Priorities for Industry Accounts at BEA, presented at the meeting of the Bureau of Economic Analysis Advisory Committee, Nov. 17, 2000. Available at

Address for Correspondence and End Notes

Brian W. Sloboda, Bureau of Transportation Statistics, U.S. Department of Transportation, 400 7th Street, SW, Room 3430, Washington, DC 20590. Email:

1. In this paper, we focus on HPTS through automobiles and ignore household production through private planes and other household-operated transportation vehicles such as boats, which should be included for a complete estimate of HPTS.

2. Similar treatment has been applied in national accounts to certain forms of nonmarket household activities (e.g., imputing and counting the value of owner-occupied residences and farm products consumed on farms). In other words, what we are doing for HPTS has a precedent in existing national accounting practice.

3. Based on the Bureau of Labor Statistics (, in 2001, the average hourly earnings for the private nonfarm industry were well above $14 while that for taxi drivers and chauffeurs were well below $10.

4. The estimate of household vmt is based on the 1995 NPTS Summary of Travel Trends and the annual growth rate of vmt by total passenger car, motorcycle, and other 2-axle, 4-wheel vehicles. Business vmt is the gap between total vmt through motor vehicles and household vmt. Caution should be taken in that vmt tells only a partial story of transportation services through motor vehicles and should not be taken as a summary measurement of transportation services through motor vehicles. For example, a given amount of vmt accomplished by a trucking company would probably indicate a significantly higher value of transportation services than that for a household driver.

5. See Fang et al. (1998), Fang et al. (2000), and USDOT BTS (1999) for details.

6. As stated in Yuskavage (2000), BEA relies on the income and expenditure approach or the gross product originating (GPO) approach for the NIPA. Although the GPO by industry and value added from the I-O Accounts are conceptually equivalent, "GPO is based primarily on BLS data for labor income and on IRS data for capital income, whereas I-O is based primarily on Census data. Differences in the industry classifications among these major data sources lead to inconsistencies. In addition, I-O value added is a residual that is critically dependent on good data for purchased goods and services by industry, which are limited. The GPO estimates of capital income depend on the assumption that corporate profits and other company-based income components can be allocated to the establishment level by industry."

7. An example of the special definitions and conventions used in the I-O Accounts, according to the Personal Consumption Expenditures monograph (USDOC BEA 1990), is "the accounting for imputed transactions," which "serves the purpose of keeping GDP, PCE, and other NIPA aggregates invariant to whether 1) housing and institutional structures and equipment are rented or owned; 2) employees are paid in cash or in kind; 3) farm products are sold or consumed on the farm; 4) saving, lending, and borrowing are direct or are intermediated; and 5) intermediated financial transactions involve an explicit or implicit service charge." More examples follow later in this paper.