5.Data Sources and Methodology

The Bureau of Economic Analysis (BEA), the U.S. Census Bureau (Census), and the Office of Management and Budget are the principal data sources. Not all data are directly taken from these sources. A number of adjustments are required due to differences in definitions, coverage, reporting periods, etc. We also fill in data gaps whenever necessary. The description of data sources and procedures used to produce the data tables are provided below.

5.1 Government Investment in Transportation

Gross government capital investment in transportation is classified as investment in infrastructure and rolling stock. The consensus is that equipment used for providing and maintaining the transportation infrastructure may be seen as a component of infrastructure despite the fact that the government invests in transportation equipment other than rolling stock (e.g., computers, communication equipment, software, office and accounting equipment, machinery, etc). It should be noted that there is no identifiable source of data that can be used for a reliable estimate of government investment in transportation equipment that is not blended with infrastructure. Besides, the available data from the Census Bureau show that government investment in transportation equipment is insignificant compared with the government investment in transportation infrastructure. Therefore, the report does not cover government investment in transportation equipment other than rolling stock.

5.1.1 Data Source and Method of Estimation for Government Investment in Rolling Stock

The government annual purchases of motor vehicles, including automobiles, trucks and buses, are estimated from BEA, National Accounts, Table for Auto Output (i.e., underlying table 7.2.5U). More specifically, government purchase of motor vehicles (excluding employee reimbursement and government investment in motor vehicles for defense) is counted as government capital outlays on rolling stock for the highway and mass transit modes.

Currently, no data are available on government purchases of other types of rolling stock like aircraft and ships. The government purchases aircrafts and ships, which are primarily used for defense purposes. These purchases are not treated as investment; instead they are counted as current consumption.

5.1.2 Data Source and Method of Estimation for Federal Government Investment in Infrastructure

New data sources and methods are used for assembling data on government investment in transportation infrastructure. The old data series of the government transportation infrastructure investments were gathered through personal communications with the Bureau of Economic Analysis (BEA). The agency compiled government transportation investment by level of government and mode of transportation through special studies. These special studies were conducted on an ad-hoc basis and their results were not published. Consequently, this data source does not allow us to regularly update the government infrastructure investment data tables and figures. It was, therefore, necessary to look for alternative data sources in order to update the government infrastructure investment data tables and figures on a regular basis (Table 1).

Federal investment in transportation infrastructure is compiled by mode, including highways, air, water, transit, and railroad. The investment data on highways and streets are directly obtained from the Bureau of Economic Analysis (BEA), Fixed Assets Table 7.5.

Federal infrastructure investment for air, water, transit and railroad is estimated based on data from the Office of Management and Budget (OMB). A number of adjustments to OMB's data are made in order to derive the final estimates of federal infrastructure investment, and make the coverage as complete as possible without double counting. The estimation process involves the following steps:

1. Identify transportation infrastructure investment related programs from the U.S. budget database.⁴
   Table 2 provides the list of federal infrastructure related programs that are included in the estimates. The list includes those programs directly funded by federal government agencies. Note that the list in table 2 is not exhaustive. There are several federal infrastructure related programs that are not included in the list, because the federal government does not make any direct payments to these programs, instead transfers funds to state and local governments.⁵
2. Calculate the non-grant portion of the outlays of those infrastructure related programs.
   The non-grant portion of these infrastructure programs represents the direct federal expenditures on infrastructure programs.
3. Make adjustments for offsetting collections whenever necessary.
   Offsetting collections are federal receipts from the public resulting from business-type activities, or transfers from other federal accounts. Such receipts are deducted from gross outlays of a receiving program in the budget of the U.S. Government, instead of combining them with federal receipts to be counted as part of the federal revenue. Thus, outlays in the budget are reported net of offsetting collections. In order to provide the actual spending of a program in this report, it is necessary to adjust for any deduction of "offsetting receipts" from program outlays. This is accomplished by examining grant and non-grant outlays of each program from the Budget Database, and by analyzing detailed program appropriations from the Appendix of the U.S Budget.
4. Convert expenditures from fiscal year to calendar year.
   The Office of Management and Budget (OMB) provides data on a fiscal year basis that starts on October 1 and ends on September 30. The estimates in this report are provided on a calendar year basis. Let CY_(t) represent the expenditure of calendar year t, CY_(t-1) the expenditure of calendar year t-1, FY_(t) the expenditure of fiscal year t, which runs from October 1 of CY_(t-1) to September 30 of CY_(t). Assuming federal outlays are evenly distributed over the four quarters of a fiscal year, we can convert expenditures from fiscal year to calendar year using the following formula:
   
   

5.1.3 Data source and Method of Estimation for State and Local Government Investment in Infrastructure

State and local government investment data for highways and streets are directly taken from BEA's Fixed Assets, Table 7.5.

State and local government infrastructure investment for air, water, and transit are based primarily on Census Bureau's annual survey of government finances. The Census Bureau reports state and local government expenditures for the construction of air, water and transit infrastructure with a two year lag. For latest years, we use the trends of state and local spending, which are obtained from Census Bureau's Construction Survey.

The data from Census Bureau's survey of government finances are provided for fiscal years ending between July 1 and June 30. For example, the FY 2002 data represent the financial activities of state and local government for fiscal year that ended between calendar year July 1 2001 and calendar year June 30 2002.

Most state governments have fiscal years beginning on July 1 and ending on June 30. However, local governments have fiscal years ending at various dates during a year. Since data in our report are provided on a calendar year basis, we need to convert Census Bureau's state and local government expenditure data from fiscal year to calendar year. For this purpose, we applied BEA's method of conversion of state and local government finance from fiscal year to calendar year. BEA uses different methods for state governments, and local governments.

For state governments, the conversion of government finance data to a calendar year basis is accomplished by a 2-year average. For example, let CY_(t) represent the expenditure of calendar year t, FY_(t) the expenditure of fiscal year t, FY_(t+1) the expenditure of fiscal year t+1, then BEA's two-year average method estimates CY_(t) as:

.

For local governments, BEA uses weights established based on periodic Census Bureau tabulations by ending month of the fiscal year. For example, let CY_(t) represent the expenditure of calendar year t, CY_(t-1) the expenditure of calendar year t-1, FY_(t) the expenditure of fiscal year t, which runs from July 1 of CY_(t-1) to June 30 of CY_(t), then BEA's formula for CY_(t) is given by: . This formula extends over three fiscal years, because fiscal years of local governments start and end at various dates during a year (see the illustration below).

In the above diagram, calendar year CY_(t) touches FY_(t), FY_(t+1), and the local governments' FY that start in August, September, October, November, December FY_(t+1) and ends in FY_(t+2). Consequently, the formula for converting fiscal year to calendar year incorporates three fiscal years.

5.1.4 Coverage of Government Investment

The coverage of government investment estimates is expanded to include investment in railroad infrastructure. This was not included in the old investment estimates.

In addition, local governments spend significant amount of money every year for providing parking lots and garages. Local government spending for the construction of these infrastructure facilities were not included in the earlier version of the report, but are included in the current report.

5.2. Business Investment in Transportation

Estimates for business investment in transportation are developed based on BEA data on capital investment by industry and by asset type. Our report classifies industries into two sectors for the purposes of measuring investment, the transportation industry, and all others, which we refer to as non-transportation industries. Six modes are specified within the transportation industry. They are: highways, air, water, mass transit, railroad, and pipelines. For the transportation sector, the following three types of assets are specified: infrastructure, rolling stock, and other. Non-transportation industries include agricultural, mining, utilities, construction, manufacturing, wholesale and retail trade, information, finance and insurance, real estate and rental and leasing, and services. For non-transportation industries, only investment in rolling stock (including motor vehicles, aircrafts, ships and boats, and railroad equipment) is identified as transportation investment.

The new private transportation infrastructure investment estimates are expanded to include private construction of parking lots and garages. This investment item was not included in the estimates of the previous report. Data on parking construction are obtained from Census Bureau's Survey of Construction for years 1993-2003. Data for years before 1993 are estimated using trends in the "other commercial construction," which has parking as a component. Investment in "other commercial construction" for years before 1993 is available from Census Bureau's Construction Survey old data series.

5.3. Household Purchase of Rolling Stock

Household purchase of rolling stock is classified into three modes: highways and streets (road), air, and water. For road transportation, household purchase of rolling stock includes automobiles, motorcycles, and bicycles. Household purchase of aircraft and boats are classified as air transportation and water transportation, respectively. The data are taken from the National Income and Product Accounts of BEA.

⁴ The entire fund for infrastructure programs may not be used for infrastructure purpose only. Instead, some of the money assigned for infrastructure programs can be used for current operational activities or for purchases of equipment. This limitation of the data should be taken into account in using the data.

⁵ For example, several highway projects financed by funds from the Highway Trust Fund are not included in the list. Almost all of the Highway Trust Fund money (with the exception of funds used for administrative and other related purposes) is transferred to states. The highway investment amounts financed by money from the Trust Fund are captured in the data for states. Federal grants to the Washington Metro Area Transit Authority for the construction of the Washington Metro Rail System is not included. These funds were used for the construction of the Franconia/Springfield, Glenmont, Mid- City, and Branch Avenue lines, and for project management, real estate acquisition, and other expenses such as contingencies and insurance.