GNP: an overview of source data and estimating methods.

Carson, Carol S.

GNP: An Overview of Source Data and Estimating Methods

GNP is the most widely used measure of the Nation's production. It is used to track the cyclical ups and downs of the economy and to monitor economic growth. In conjunction with measures of labor and capital input, it provides indicators of the Nation's productivity. Thus, GNP is a primary focus in the analysis of the current economic situation, in macroeconomic forecasting, and in policy formulation.

The usefulness of GNP stems in part from its being the keystone of a set of measures that provide answers to two questions: First, what is the output of the economy--its size, its composition, and its use? Second, what is the economic process or mechanism by which output is produced and distributed? These measures are in the framework of a set of accounts--the national income and product accounts (NIPA's)--that show production, distribution, consumption, and saving. The summary NIPA system is shown on pages 16 and 17 of this issue of the SURVEY.

The first account in the summary system, the national income and product account, shows GNP measured in two ways (table 1). On the right side, GNP is measured as the sum of goods and services sold to final users. On the left side, GNP is measured as the sum of income payments and other costs. The right side, which corresponds to table 1.1. of the set of NIPA tables, shows personal consumption expenditures for goods and services and the other "product' components. The left side shows the compensation of employees and the other "income' components; most of these entries appear as components of national income in NIPA table 1.14; the remainder appear in table 1.9, which shows the relation among major NIPA income and product aggregates.

As explained in "An Introduction to National Economic Accounting,' the national income and product account is what is referred to as a "production account.'[17]1 Specifically, it is the production account for the Nation. Although most of the Nation's production takes place in the business sector, production also takes place in the other three sectors of the economy--households, government, and foreign. Thus, this account can be viewed as the consolidation of the production accounts for all economic transactors, based either on actual production accounts or those that can be assumed to exist. The national income and product account also serves as the appropriation account for the business sector; that is, it records the business sector's current income, current outlays, and saving. The other accounts in the summary system are appropriation accounts for the three other sectors and a consolidated saving-investment account for the domestic sectors.

1. Bracketed numbers refer to items in the second part of this article.

The first part of this article summarizes the source data and estimating methods used to prepare the current- and constant-dollar estimates of GNP. This overview of GNP complements the more detailed information that is becoming available in BEA's series of papers describing the conceptual framework of the NIPA's and the methodology used in preparing the estimates. The second part provides a directory of the information about GNP that has appeared in recent years in the SURVEY OF CURRENT BUSINESS and other BEA publications. GNP and the other entries in the national income and product account are defined next.

Measures of production.--As indicated, the total on both sides of the national income and product account is GNP. Gross national product, to use the words that underlie the familiar acronym, is defined as the market value of the goods and services produced by labor and property supplied by residents of the United States. The product side of the account shows that GNP is the sum of purchases of goods and services by persons and government, gross private domestic investment (including the change in business inventories), and net exports (exports less imports). This sum is the market value of final sales and business inventory change. To avoid double counting, the summing does not include business purchases of goods and services on current account--for example, business purchases of paper to make books; their value is reflected in the value of final sales and business inventory change. The investment component of GNP is measured before deduction of charges for consumption of fixed capital.

The income side shows charges against GNP (line 25), the costs incurred and the profits earned in the production of GNP. This sum equals GNP, except for the statistical discrepancy. The factor charges--compensation of employees, proprietors' income, rental income of persons, corporate profits, and net interest--represent the incomes of the factors of production (labor and property). These factor incomes are measured before deduction of taxes on those incomes and after deduction of depreciation and other allowances for capital consumed in production. The total of these incomes is a measure of production called national income. Three nonfactor charges--business transfer payments, indirect business taxes, and the current surplus of government enterprises--are added to national income and one--subsidies--is subtracted to yield charges against net national product, yet another measure of production. Charges for the consumption of fixed capital are added to charges against net national product to yield charges against GNP.

The four measures of production that have been mentioned so far differ from each other because of distinctions that are made between measures at factor cost and at market value (that is, including nonfactor charges as well as factor costs) and between measures that are gross and net of capital consumption. GNP and charges against GNP are gross market value measures. National income (line 19), the income that originates in the production of the goods and services attributable to labor and property supplied by residents of the United States, is a net factor cost measure (line 19). Charges against net national product (line 23), the income that originates in, plus the nonfactor charges incurred in, the production of the goods and services attributable to labor and property supplied by residents of the United States, is a net market value measure.

One further basic distinction can be made in defining the measures of production. This distinction is between national measures and domestic measures. As indicated by its name, the national income and product account presents measures on a national basis. National measures relate to the ownership of the factors of production; they refer to production attributable to labor and property supplied by residents of a country. Domestic measures relate to the physical location of the factors of production; they refer to production attributable to all labor and property located in a country. The national measures differ from the domestic measures by the net inflow--that is, inflow less outflow--of labor and property incomes from abroad.

The choice of a measure of production from this array depends on the use to which it will be put. For example, national income is often used in studies dealing with the allocation of factors of production to various uses. A market price measure, such as GNP, is usually preferred for studies of economic behavior and welfare, because market price is the basis for choice among alternative products. Gross domestic product, rather than GNP, is the most widely used measure of production in other countries, and thus it is often used in international comparisons.

Components.--The income and product components of GNP are defined in the box on pp. 104-5. They are in the order of the line numbers in the national income and product account, starting on the income side and proceeding to the product side. These components are often defined by reference to the purchaser of the goods and services or to the payor or recipient of the income. "Business,' in the main, refers to corporate and noncorporate private entities organized for profit. Mutual financial institutions, private noninsured pension funds, cooperatives, nonprofit organizations that primarily serve business, Federal Reserve banks, and federally sponsored credit agencies are also included as part of business. Government enterprises (such as the Commodity Credit Corporation and Stateowned liquor stores) are treated in some respects like private business and in other respects like other government agencies. "Persons' refers to individuals, nonprofit institutions serving individuals, private noninsured welfare funds, and private trust funds. ("Households and institutions' excludes the two kinds of funds.) "Government' refers to all Federal, State, and local government agencies. "Foreigners' refers to foreign residents who are transactors with U.S. residents; "rest of the world' is used with the same meaning.

Part I. Source Data and Estimating Methods

This part summarizes the source data and estimating methods used to prepare GNP. Such a summary encourages appropriate use of the estimates by making the content of the income and product components more concrete than did the definitions. For example, the definition of fixed investment refers to capital goods "purchased;' the summary methodology clarifies that the estimate for structures is of the value put in place in a time period, which would differ from the full purchase price if the construction is underway in several time periods. Such a summary also makes explicit that the component estimates rely in varying degrees on proxies and judgment--a reminder that, although the estimates are useful analytical tools, they cannot be precision instruments.

Source data are the facts and figures BEA uses to prepare the estimates, and estimating methods are the steps BEA takes to conform the source data to the concepts and definitions of the NIPA's and to fill in gaps in data coverage. Taken together, source data and estimating methods are often referred to as "methodology.' The summary in this article highlights features of the methodology that serve the purposes just mentioned; they throw light on the content of the components and provide indications of reliability.

The methodology described--first for current-dollar estimates and then for constant-dollar estimates--is that in use as of the July 1987 annual revision. Methodology is not fixed once and for all. First, source data emerge and disappear, and BEA adapts the estimating methods accordingly. Second, often independently of circumstances that lead to these adaptations, BEA improves the estimating methods. In recent years, a number of improvements were associated with the expanding use of computers; most obvious perhaps is the ability to handle a finer level of estimating detail in the allotted timeframe. Finally, the economy changes, and BEA updates the methodology to ensure that the estimates continue to provide a reliable picture of the transactions and transactors in the economy. One notable instance of methodological change--the improvements in the constant-dollar estimates and associated price measures over the last decade--is evident from the directory of information about the estimates, in the second part of the article. In this instance, the impetus was the inflation that began in the early 1970's, and the improvements involved developing new source data, refining other source data, and adapting the estimating methods to include more sophisticated treatments and to handle more detail.

Current-dollar estimates

Preview of source data and the estimating cycle.--With few exceptions, the data used in preparing the estimates are collected for purposes other than the preparation of the income and product estimates. Data collected by Federal Government agencies provide the backbone of the estimates, although they are supplemented by data from trade associations, businesses, organizations, and other private sources. The government data come from a number of agencies, mainly Commerce Department's Census Bureau, Treasury Department's Internal Revenue Service, Labor Department's Bureau of Labor Statistics, Office of Management and Budget, and Agriculture Department. Some government-collected data, referred to as "administrative' data, are byproducts of government functions such as welfare and social security programs, tax collection, defense, and regulation. Nonadministrative data, sometimes referred to as "general purpose' or "statistical' data, include the periodic economic and population censuses and a wide range of sample surveys, such as those that collect data on manufacturing activity, corporate profits, and the labor force status of households. Of the relatively few items for which BEA itself collects data used to estimate the NIPA's, most refer to international transactions. These include investment by foreigners in the United States and by U.S. residents in foreign countries.

The article in this issue about the annual NIPA revision highlights an important characterstic of the estimates: The source data on which they depend improve in quantity and quality as the estimates progress through the estimating cycle. A cycle usually takes 5 years and runs from three estimates of the current quarter to three (usually) annual revisions and, finally, to a comprehensive revision (often referred to as a "benchmark' revision). For the preliminary estimate of GNP for the current quarter, which is released about 3 weeks after the end of that quarter, source data provide less than full coverage and usually are subject to revision. The data available for this estimate are series collected monthly. One important series, retail sales, is released early in the month following collection, and the preliminary GNP estimate is timed to incorporate it. Thus, the preliminary estimate incorporates 3 months of data for that series. The preliminary estimate also incorporates 3 months of data for auto and truck sales and for employment, hours, and earnings. For several other important monthly series, 2 months of data are available.

In the 2 months immediately following the preliminary estimate, BEA prepares two revised estimates of GNP for the current quarter. These estimates, referred to as the "first revision' and "second revision,' each incorporate both revised and previously unavailable monthly data. As well, they incorporate three major sets of quarterly data. One set is the data on domestic corporate profits. Except for the fourth quarter, a preliminary corporate profits estimate is released with the first revision of GNP and a revised estimate with the second revision of GNP; for the fourth quarter, the corporate profits schedule lags 1 month. The second set is the data on foreign transactions from the BEA balance of payments accounts. With the exception of merchandise trade data, most of the source data underlying these accounts, including the BEA surveys of direct investment, are quarterly. Data from the balance of payments are incorporated in the second revision. The third set is the data on assets from the flow of funds accounts prepared by the Federal Reserve Board, also incorporated in the second revision.

Most quarterly and monthly source data are based on samples. The sample-based data include the monthly Census Bureau reports on shipments or sales and inventories of manufacturing, wholesale trade, and retail trade (the sales series mentioned earlier); the corporate profits data for manufacturing, mining, and trade in the Quarterly Financial Report prepared by the Census Bureau; the monthly data on employment, hours, and earnings compiled from reports of nonfarm establishments by the Bureau of Labor Statisties and the monthly data on employment collected from households by the Census Bureau in the current population survey, both published by the Bureau of Labor Statistics in Employment and Earnings; and the monthly data reported by the Census Bureau in Value of New Construction Put in Place.

Several sets of source data are not samples and provide complete or nearly complete coverage. One monthly set of data, that reported by the Census Bureau in Highlights of U.S. Export and Import Trade, is from documentation intended to cover nearly all goods as they enter or leave the country. (The monthly trade data and the monthly construction data that are used for current quarterly estimates are not replaced by annual data based on a more inclusive collection system and are used for annual estimates as well.) Two other sets that provide complete coverage are the Federal receipts and expenditures in the Treasury Department report referred to as the Monthly Treasury Statement and the unit sales of autos and trucks from a trade source.

Ordinarily, the GNP estimates for the most recent calendar year and usually the 2 preceding years are revised each year in July. These annual revisions are timed to incorporate annual source data, although they also incorporate quarterly data released too late to be used in the current quarterly estimates. In addition, the seasonal adjustments applied to the source data are updated. (See the box on page 115 for more about seasonal adjustment.)

Annual data are usually based on samples. The schedule of one of the sample-based sets of annual data--data related to business income from tabulations of tax returns --is the major reason for the third annual revision. This data is compiled by the Internal Revenue Service and published in its Statistics of Income series. Data on sole proprietors and partnerships become available for the second annual revision, and data on corporations become available for the third.

The sample-based data also include the annual counterparts of the data on shipments or sales and inventories mentioned earlier (reported by the Census Bureau in the Annual Survey of Manufactures, the Annual Retail Trade Report, and the Annual Wholesale Trade Report); receipts of service establishments reported by the Census Bureau in the Service Annual Survey; data on agriculture--production, inventories, costs, and more--collected in several surveys by the Department of Agriculture; receipts and expenditures of State and local governments reported by the Census Bureau in Governmental Finances; and data collected by the Census Bureau in the biennial housing survey.

The annual revisions also incorporate data compiled by the Bureau of Labor Statistics on wages and salaries of employees covered by State unemployment insurance. These data, published in Employment and Wages, have provided nearly complete coverage of private nonfarm and of State and local government employees in recent years. (Data from this source and from the Internal Revenue Service tabulations mentioned above are not replaced by data based on a more inclusive collection system and are used for comprehensive revisions as well as for annual revisions.)

Comprehensive revisions are carried out about every 5 years. The comprehensive revisions--which also incorporate definitional and classificational changes--are timed to incorporate the quinquennial economic censuses; the year of the economic censuses is referred to as the "benchmark year.' The source data available for the comprehensive revisions provide the most complete coverage. Examples are the decennial census of population, the decennial census of housing (and its supplementary survey of residential finance), the quinquennial census of governments, and the quinquennial economic censuses, all collected by the Census Bureau. The economic censuses, currently conducted in years ending in "2' and "7,' provide extensive data on manufactures, wholesale trade, retail trade, service industries, construction industries, agriculture, transportation, and mineral industries. (An alternative view of the use of source data is that these censuses are the major sources used to prepare BEA's input-output tables and that the input-output tables, in turn, are the single most important source for the comprehensive revisions of the NIPA's. The input-output tables provide the level for the product--that is, expenditures--components of GNP in the benchmark year.)

The data from these decennial and quinquennial censuses as well as data (such as from the tabulations of business tax returns and the unemployment insurance system) that are not replaced--that is, are considered final--determine the levels of the component estimates. Other source data are used to interpolate and extrapolate; as will be explained more fully in the next section, these source data provide the basis for the annual and quarterly changes in the level. Annual data are usually used to interpolate between and extrapolate from the final levels. Quarterly data are used to interpolate between and extrapolate from annual levels.

Notes to the sources and methods table.--Table 2 identifies the principal source data--those previewed as well as many others--and estimating methods used to prepare the current-dollar annual and quarterly estimates of GNP components. The components are as shown in table 1, starting on the income side and proceeding to the product side. The subcomponents in table 2, with their 1986 dollar values, are grouped according to the methodology used to prepare them.

The column in table 2 for annual estimates covers the several annual estimates in the estimating cycle; major differences in methodology as the estimates move through the annual revisions to a comprehensive revision are few enough to condense into a table. For example, for most goods in personal consumption expenditures (the first item on the product side), the table indicates one methodology for benchmark years and another for all other years.

The column for the quarterly estimates is a condensation in several respects. First, it refers to the preliminary estimates for the current quarter. That one estimate, rather than all of the current quarterly estimates, is described because more attention focuses on the "first look' at the quarter. Second, even for the preliminary estimate, the column does not detail how many months of data are available nor whether the data are subject to revision by the source agency. Thus, the benefit of condensation is at the cost of not detailing the tradeoff between the timeliness of the preliminary estimate and the improved quantity and quality of source data available in the 2 following months.2

2. The methodology for personal consumption expenditures illustrates the sequential introduction of source data in the current estimates. For the preliminary estimates, the key data available are: 3 months of retail sales (of which 2 months are subject to revision by the source agency); 3 months of unit sales of new motor vehicles, 1 month of information with which to allocate the unit sales among consumers and other purchasers, and 2 or 3 months of average list prices (which are subject to modification as more detailed information on models of cars sold, and thus actual retail prices, becomes available); and 1 or 2 months of data for services amounting to three-fifths of total services. The estimates for most of the remainder of personal consumption expenditures are extrapolations based either on related indicator series or on judgmental trends. For the first revision, revised retail sales for 2 months become available. For the second revision, information on sales of used cars and more data on several services--hospitals, electricity, natural gas, telephones, airline transportation, foreign travel, and insurance other than life insurance-- become available.

The "Business Situation' in the SURVEY provides a footnote at the time of the preliminary estimates of GNP (January, April, July, and October) listing major source data, including the number of months for which they are available, for that GNP estimate. Notes to a table in the "Business Situation' in the 2 following months list source data, including revisions to earlier data, that subsequently became available and were incorporated in that estimate.

The information in the column for the quarterly estimates is generic; it fits any preliminary estimate for the current quarter. For users who want to track the actual data and resulting estimates, BEA prepares "Key Source Data and Projections for National Income and Product Estimates.' This table, provided at the time of the preliminary estimates, shows the monthly figure for key source data that are available (and indicates whether the data are subject to revision by the source agency) and BEA's projected figures for data that are not available [40]. BEA is reformatting this table and plans to make it more accessible by putting it on the Commerce Department's Economic Bulletin Board.

Table 2 lists source data referring to a variety of different economic measures--wages and salaries, premiums, expenses, interest rates, mortgage debt, tax collections, unit sales, housing stock, employment, and average price, to name a few. For most components, the source data are "value data;' that is, they embody both the quantity and price dimensions that are required for current-dollar estimates. In these cases, the methodology indicated in table 2 is the adjustment of the value data to derive estimates consistent with NIPA definitions and coverage. The descriptions of the adjustments to derive domestic corporate profits and change in business inventories are examples. Three other methods are described more fully in the next section.

When value data are not used in preparing an estimate, the table indicates the combination of data with separate quantity and price dimensions that is used to derive the required value estimate (as well as indicating any major adjustments needed to derive estimates consistent with NIPA definitions and coverage). On the product side, a "physical quantity times price' method is used for several components. For example, the estimate for new autos is prepared as unit sales times average list price. Both the income and product sides refer to an "employment times hours times average hourly earnings' method and to several forms of a "stock of assets/liabilities times an interest rate' method.

As mentioned earlier, some of the source data shown in table 2 for the annual estimate are used to interpolate and extrapolate the levels established by source data that are viewed as final, and all of the source data shown for the preliminary quarterly estimates are used to extrapolate the level of the preceding quarter. In addition to using indicator series, as is the case when source data are listed in the table, extrapolation and interpolation may use trends, as is the case when "judgmental trend' is listed in the table.

Extrapolation is a method of extending estimates from one period forward (or backward) in time to other periods. In simple terms, extrapolation applies a percent change-- either the percent change in the indicator series or the percent change in the trend--to the level of the preceding period. Interpolation is a method of filling in estimates between two periods. Interpolation applies a more complex mathematical formula--there are several in use--to preserve the pattern of the indicator series consistent with the level of the source data viewed as final. (An explanation of interpolation and extrapolation, using simple numerical examples, in the context of BEA's regional estimates is in [21], pp. T-34 and T-35.)

Illustrative methods.--Table 2 refers to three methods-- commodity flow, retail control, and perpetual inventory-- used for specific components. These methods are described briefly in the following paragraphs to illustrate the work of conforming source data to the concepts and definitions of the NIPA's and filling in gaps in data coverage.

The commodity-flow method was originally developed in the 1930's as a way to obtain the value of consumers' purchases of goods (that is, commodities) or of producers' purchases of durable equipment. A generalization of that method is now applied for these components within the input-output framework that is used to establish GNP in benchmark years. The steps and information used in benchmark years for producers' durable equipment except autos and scrap are summarized in the tabulation.

The commodity-flow method is also used for producers' durable equipment in nonbenchmark years, but it is implemented in an abbreviated form. It uses less commodity detail--starting with shipments for about 300 commodity classes from the Census Bureau annual survey of manufactures --and assumes the same distribution in steps 3-6 as in the last benchmark. (See [9] for a description of the method as applied to computers.) An even more abbreviated commodity-flow method is used for current quarterly estimates.

The strength of the commodity-flow method is that it draws on the very detailed commodity classification and comprehensive coverage of the census of manufactures for benchmark years. To go from domestic manufacturer's shipments to business purchases, it relies on data drawn from several other comprehensive, but not necessarily consistent, sets of source data. As applied in its abbreviated form, when the data for steps 3-6 are not available, the method relies on the assumption that the relationships do not change rapidly. The assumption that, for each type of equipment, the share of the total going to intermediate purchases, persons, and governments is stable between benchmark years is an important limitation.

The retail-control method, which is used to estimate over one-third of the value of personal consumption expenditures, is more direct than the commodity-flow method. The method provides, first, the indicator series used in extrapolating and interpolating the total of most goods and, second, provides the "control' total to which type-of-good components in this group are made to sum. Specifically, it is used to prepare estimates for all goods except motor vehicles, gasoline and oil, food furnished employees, expenditures abroad by U.S. residents, and personal remittances in kind to foreigners, each of which is shown separately in table 2 (and the very small amounts of food produced and consumed on farms and standard clothing issued to military personnel, which are not shown separately). The steps and information used for each current month's estimate are explained in the tabulation. For current quarters, the monthly estimates are summed to quarterly estimates. The same steps are used with comparable information for annual estimates except in benchmark years.

The retail-control method makes it possible to use retail sales, available only by kind of business except at 5-year intervals, to provide type-of-good detail. Further, its feasibility on a monthly basis is the foundation for the monthly estimates of personal consumption expenditures. These monthly estimates, in combination with monthly estimates of personal income, provide timely information for the analysis of the economic activities of one of the major sectors in the economy. The basic limitation of the method as an extrapolator is that it cannot take into account changes in the relative importance of business purchases at retail stores, services sold at retail stores, or personal purchases outside of retail stores. In addition, use of the method to provide the control to which components sum relies on the general assumption that the types of goods purchased in the various kinds of retail stores do not change rapidly.

The perpetual-inventory method is used to derive estimates of fixed capital stock. The stock estimates, in turn, form the basis for the estimates of capital consumption allowances with capital consumption adjustment (often referred to as "economic' depreciation) and for the estimates of the capital consumption adjustment shown with proprietors' income, rental income of persons, and corporate profits. The perpetual-inventory method is based on cumulating past investment flows and deducting the value of discarded assets; it is an alternative to direct measurement of the capital stock, which is seldom statistically feasible on a comprehensive basis. The steps and information used in preparing BEA's estimates of the capital stock are summarized in the tabulation.

One of the main advantages of obtaining the stock estimates in this way is that, for the most part, comprehensive, detailed, and relatively reliable estimates of investment flows--with which the method starts--are available within the NIPA's. By using these flows, the resulting stock and depreciation estimates are consistent conceptually and statistically with the NIPA's. In addition, the perpetual-inventory method provides alternative valuations of the capital stock--at historical cost (that is, when an asset is valued at the price at which it was purchased new), at constant cost (that is, when an asset is valued at the price that would have been paid for it had it been purchased in the base period), and at current cost (that is, when an asset is valued at the price that would have been paid for it had it been purchased in the period to which the stock estimates refer). One of the disadvantages of this method is that it relies on information on the service lives of various kinds of assets; accurate and up-to-date information of this kind is difficult to obtain.

Observations.--Although the main purpose of the sources and methods table is to provide a convenient reference, the overview it provides serves as the basis for several broad observations about methodology.

First, on the product side, separate methodologies are often shown for benchmark and other years. On the income side, the distinction between benchmark and other years seldom appears. For many income-side components, the same methodology is used each year. However, for several --notably corporate profits and net interest--the final source data do not become available for several years so that interim methologies must be used until the final one can be substituted.

Second, product-side estimates are largely based on non-administrative data. Census Bureau censuses and surveys dominate; for the annual estimates, they account for about two-thirds of the product-side total. Income-side estimates are largely based on administrative data. Tabulations from the unemployment insurance program and from business income tax returns dominate; for the annual estimates, they account for about two-thirds of the income-side total.

Third, the product-side estimates provide detail by type of good or service. Identifying each sector's purchases is fundamental to the structure of the national income and product account, and that requirement dictates the use of source data--notably manufacturers' shipments as used in the commodity-flow method--that also provides detail for many components. (Also, as indicated in the next section, detail is crucial to the preparation of constant-dollar estimates.) The income-side estimates provide detail by industry, by legal form of organization (for example, corporate and noncorporate business), or both.

Finally, the estimates of personal consumption expenditures provide an interesting contrast between the methods of deriving totals and component detail. When the retail-control method is used, the total of expenditures for goods prepared in that way is estimated with more certainty than is the component detail, which is derived by allocating the total using percentages determined at 5-year intervals. In contrast, the total of expenditures on services is pieced together from component detail. The estimating procedure is simpler in that the consumer of a service often purchases it directly from the producer and many types of services--haircuts, for example--are purchased only by persons. However, any uncertainty about the source data that provide the detail carries through to the total of expenditures for services.

Constant-dollar estimates and associated price indexes

The principle underlying constant-dollar estimates is simple: Value each component at its price in a base period. (See the box on page 115 for more about the base period.) This approach, rather than the approach of adding the various physical quantities (the approach perhaps suggested by the term "real' that is often used in referring to constant-dollar estimates), is taken because "you can't add apples and oranges.' Constant-dollar values provide the means of aggregation.

This principle is implemented for the product-side components of GNP by one of three methods, depending on the availability of source data. The method used for by far the largest part of GNP is deflation. In fact, deflation is so widely used that the term is often used to describe the preparation of all constant-dollar estimates. In deflation, constant-dollar estimates are obtained by dividing the most detailed current-dollar components by appropriate price indexes with the base period--at present, the year 1982--equal to 100.

The other methods, quantity extrapolation and direct base-year valuation, are similar in that they both use quantity data. For quantity extrapolation, constant-dollar estimates are obtained by extrapolating the current-dollar estimates from the base year to the current period by quantity data. For direct valuation, constant-dollar estimates are obtained by multiplying base-year prices by quantity data for the current period.

In preparing constant-dollar estimates, detail is crucial because the greater the detail the more closely the results approximate the ideal of pricing each good or service purchased in the current period with its base-year price. A number of the improvements in methodology for the constant-dollar estimates have focused on carrying out the calculations at a more detailed level. For example, until the mid-1970's, expenditures on food consumed at home had been deflated by the index for food at home from the Consumer Price Index, in which food prices are weighted by the composition of expenditures in a base period. The improvement was to disaggregate these expenditures and deflate 16 separate categories with the appropriate component of the index. Thus, if there are shifts in expenditures between the base year and the current year--for example, from meats to vegetables--the more detailed deflation captures the shifts and provides a more accurate measure of changes in constant-dollar expenditures on food. At present, roughly 600 categories of goods and services are deflated separately for the quarterly and regular annual estimates.

Table 3 summarizes the methodology for the preparation of constant-dollar estimates of GNP and its product-side components, showing which of the three methods is used and indicating the source data with which it is implemented.3 The subcomponents are as shown in table 2, except where more detail is needed to highlight differences in methodology for constant-dollar estimates. For example, more detail is provided for Federal purchases because distinctly different source data are used for the deflation of national defense and nondefense purchases. For this table, the distinction between annual and quarterly methodology is far less important than it was for the current-dollar methodology, and major differences between the annual and quarterly source data are noted within the individual entries.

3. With few exceptions, BEA does not prepare constant-dollar estimates of income measures because price indexes cannot be associated with them, as they can with product measures. Two exceptions are disposable personal income and the profits measures presented in BEA's Business Conditions Digest. In both cases, the series are adjusted for price change by reference to prices of the goods and services on which the income is spent. BEA prepares constant-dollar net national product and national income by preparing constant-dolalr estimates of capital consumption allowances with capital consumption adjustment and of the nonfactor charges and then subtracting them from constant-dollar GNP.

Table 3 shows that components of the Consumer Price Index and the Producer Price Index, prepared monthly by the Bureau of Labor Statistics, are used to deflate many components, which together account for well over three-fourths of GNP. Other indexes used in deflation are mainly for services (indexes of earnings and expense), structures (various cost indexes and a Census Bureau index for new houses), merchandise exports and imports (Bureau of Labor Statistics and Census Bureau indexes), and national defense (primarily BEA indexes). Quantity extrapolation is used for a few service subcomponents in personal consumption expenditures, mining shafts and wells, imputed interest and financial service charges, and compensation of employees in government purchases. Direct valuation of quantities is used for used autos and inventories of agricultural commodities held by farmers and the Federal Government.

The implicit price deflator and price indexes

The preparation of current- and constant-dollar estimates yields a byproducts--the implicit price deflator. It is derived as a current-dollar measure divided by a constant-dollar measure, multiplied by 100. Technically, the implicit price deflator for GNP or a component of GNP is an average of the indexes of prices of all the goods and services that make up GNP or the component, weighted by the constant-dollar composition of GNP or the component in the current period. Thus, changes in the implicit price deflator reflect not only changes in prices but also any shift in the composition of GNP or the component.

A fixed-weighted price index also is an average of the indexes of prices of all the goods and services that make up GNP or the component. In contrast to the implicit price deflator, the index is weighted by the composition of GNP or the component at a fixed point selected as the base period. Such indexes measure the changes in the price of a fixed market basket; thus, they measure only price change. Chain price indexes, from one period to the next, also do not reflect shifts in the composition. In the 1970's, it became especially useful to distinguish quarter-to-quarter shifts in the composition of GNP from price changes because at times each was large. Since then, BEA has increasingly featured the GNP fixed-weighted and chain price indexes as appropriate measures of price change.

Part II. A Directory to Information about GNP

In recent years, a number of BEA publications have provided information about GNP and its income and product components. The following is an annotated list of those publications, numbered with the most recent publication first and going back to the mid-1970's. Among the items listed are the methodology papers mentioned earlier and articles describing the comprehensive and annual revisions. The methodology papers explain the conceptual framework of the NIPA's and the source data and estimating methods used to prepare the estimates. One of the papers provides an introduction to national economic accounting; it is listed as item 17. Two others, on foreign transactions and corporate profits, have been published; they are listed as items 3 and 15, respectively. During the period covered by the list, comprehensive revisions were released in 1985, 1980, and 1976; the articles describing them are listed as items 13, 34, and 47. Annual revisions were released in most years (except before or after comprehensive revisions); the articles describing them are listed as items 1, 7, 19, 24, 29, 37, 41, 43, 45, and 53. A number of the other publications in the list, although more general or less directly related, contain useful information about GNP and its components. These include publications about reliability of the estimates (items 10 and 54), capital stock (items 2 and 14), regional income estimates (item 21), and the underground economy (items 18 and 20).

Table 4 is designed to serve as a directory to the information in these publications. For GNP and several other measures of production and for each income and product component, the first two columns indicate the publications that have included descriptions of methodology and the second two columns indicate publications that have discussed changes in definition and classification and changes in methodology. The publications about the methodology for constant-dollar estimates or about price measurement are marked with an asterisk. The items indicated for the production measures tend to be overviews; those for components, more specific.

1. "The U.S. National Income and Product Accounts: Revised Estimates, 1984-86, and First and Second Quarters 1987.' July 1987 SURVEY. Presents a regular annual revision. Describes methodological changes affecting personal consumption expenditures (different source data for quarterly estimates of gasoline, first-time estimates of video cassette rentals, and, for current-dollar estimates, an adjustment for quality change in used car purchases); producers' durable equipment (more detailed treatment of exports and imports in the abbreviated commodity-flow method and, for current-dollar estimates, an adjustment for quality change in used car transactions); net exports (added coverage of commissions on securities transactions and of medical services); and deflation procedures. The changes in deflation procedures affected personal consumption expenditures (a different price index for new trucks); residential investment (a modification of the price index for new houses); and personal consumption expenditures, producers' durable equipment, net exports, and government purchases (inclusion of prices for personal computers).

2. Fixed Reproducible Tangible Wealth in the United States, 1925-85. 1987 (GPO Stock No. 003-010-00177-1, price $18.00). Includes a detailed description of the investment flows used to implement the perpetual-inventory method of preparing the capital stock estimates and of the derivation of the associated estimates of depreciation. Also describes the recently developed estimates of fixed private capital by industry.

3. Foreign Transactions in the National Income and Product Accounts. BEA Methodology Paper No. 3, 1987 (GPO Stock No. 003-010-00178-0, price $2.75). Describes the concepts, sources, and methods of the NIPA foreign transactions components, including the net exports component of GNP. In particular, describes the conceptual and statistical relationship between the NIPA foreign transactions and the balance of payments accounts. See also item 5.

4. "Gross Product by Industry, 1986.' April 1987 SURVEY. Includes a summary of sources and methods for the estimates of GNP by industry, for which the current-dollar estimates are prepared as distributions by industry of the income-side components of the national income and product account.

5. "Foreign Transactions in the National Income and Product Accounts: An Overview.' November 1986 SURVEY. Presents the conceptual basis and framework of foreign transactions in the NIPA's, describes the presentation of the estimates, and summarizes the sources and methods used to prepare them. See also item 3.

6. The National Income and Product Accounts of the United States, 1929-82: Statistical Tables, 1986 (GPO Stock No. 003-010-00174-7, price $23.00). Includes a full set of definitions of NIPA aggregates and components, a brief discussion of constant-dollar estimates and price indexes, and descriptions of the classification of production by sector, legal form of organization, and industry.

7. "The U.S. National Income and Product Accounts: Revised Estimates, 1983-85 and First Quarter 1986.' July 1986 SURVEY. Presents a regular annual revision. Describes methodological changes affecting State and local indirect business tax and nontax accruals (to include fines levied under Federal oil price regulations and paid to States as a nontax); corporate profits (adjustments for profits reported on the completed-contract method of accounting, for reversions of defined benefit pension plans, for employer contributions to tax-credit stock ownership plans, and for the amount of fines paid); farm proprietors' income and farm corporate profits (incorporating a defaulters' gain); and deflation procedures. The changes in deflation procedures affected personal consumption expenditures (for airline transportation, a BEA discount-adjusted price index; for computers, a BEA price index that incorporates detailed information on changes in prices and characteristics of computers) and producers' durable equipment (eliminating the lagging of price indexes previously needed to put them on a delivery basis). In describing revisions, presents summary methodologies for services in personal consumption expenditures and for net interest and personal interest income.

8. "Federal Farm Programs for 1986-90.' April 1986 SURVEY. Includes a statement of the NIPA treatment of the transactions of the Commodity Credit Corporation.

9. "Improved Deflation of Purchases of Computers.' March 1986 SURVEY. Presents the quality-adjusted deflators constructed by BEA for computers included in producers' durable equipment, exports and imports, and government purchases, primarily from price indexes developed by IBM Corporation. Discusses the selection of the index most appropriate for deflation of computers in GNP, the construction of the new deflators, and their use in the estimates presented in the comprehensive revision released in 1985. See also "The Economic Interpretation of Hedonic Methods' and "Quality-Adjusted Price Indexes for Computer Processors and Selected Peripheral Equipment' in the January 1986 SURVEY and "Corrections to the Estimates of Purchases of Computers' in the March 1986 SURVEY.

10. The Use of National Income and Product Accounts for Public Policy: Our Successes and Failures. BEA Staff Paper No. 43, 1986 (GPO Stock No. 003-010-00144-5; price $3.75). Reviews the "accuracy' of NIPA estimates, using the size of the revisions to GNP as an indicator, and users' recommendations over the last 30 years for new and improved series, schedules of release, changes in concept and structure, etc.

11. "Revised Estimates of the National Income and Product Accounts of the United States, 1929-85: An Introduction.' December 1985 SURVEY. Presents a comprehensive revision; see item 13. Includes notes on the effect of shifting the base period and on the revision of producers' durable equipment. Presents a table indicating new NIPA tables and items.

12. "A Note on Merchandise Trade Data.' October 1985 SURVEY. Describes the changes in the Census Bureau monthly data on merchandise trade that are used as source data for the NIPA quarterly estimates. (Further changes were made in 1987; they were noted in the announcement of the consequent rescheduling of BEA's release of several estimates in the February 1987 SURVEY.)

13. "An Advance Overview of the Comprehensive Revision of the National Income and Product Accounts.' October 1985 SURVEY. Describes definitional and classificational changes: Two relate to classification of certain business expenditures--replacement railroad track and major replacements to residential structures--as investment, three relate to Federal Government employee benefit programs, three relate to foreign transactions (one of which is described in item 11), and two relate to government assistance programs. Describes three methodological changes affecting a number of components--improved adjustments for misreporting on tax returns ("underground economy adjustments'--see item 20), improved estimates of capital consumption allowances (based on new work on capital stock estimates), and exclusion of the Commonwealth of Puerto Rico and U.S. territories. Also describes a change affecting rental income of persons (improved accounting for expenses of homeownership); a shift from 1972 to 1982 in the base year for the calculation of constant-dollar estimates and price indexes; changes in deflation procedures; and a number of other methodological changes. The changes in deflation procedures affected the several components that include for computers (a BEA price index that incorporates detailed information on changes in prices and characteristics of computers--see item 9), non-residential buildings, residential additions and alterations (see item 24), Commodity Credit Corporation purchases, national defense purchases, and net exports.

14. "Fixed Private Capital in the United States: Revised Estimates, 1925-81, and Estimates by Industry, 1947-81.' July 1985 SURVEY. Includes a description of the investment flows used to implement the perpetual-inventory method of preparing the capital stock estimates and of the derivation of the associated estimates of depreciation. (See item 2 for an updated, fuller description.)

15. Corporate Profits: Profits Before Tax, Profits Tax Liabilities, and Dividends. BEA Methodology Paper No. 2, 1985 (GPO Stock No. 003-010-00143-7, price $2.50) Describes the concepts, sources, and methods of the corporate profits components of the NIPA's.

16. "A Note on the Implicit Price Deflator.' May 1985 SURVEY. Describes the implicit price deflator for GNP and contrasts it with the GNP fixed-weighted and chain price indexes. (A more technical explanation appears in "Alternative Measures of Price Change for GNP' in the March 1969 SURVEY.)

17. Introduction to National Economic Accounting. BEA Methodology Paper No. 1, 1985 (GPO Stock No. 003-010-00158-5, price $1.00). Reprint of an article that appeared in the March 1985 SURVEY. Places the U.S. NIPA's within the larger framework of national economic accounting. Shows the step-by-step derivation of a general national economic accounting system from the conventional accounting statements used by business and government and inferred for other transactors. Also shows how the income and product accounts, the capital finance accounts, and input-output accounts--the major branches of national economic accounting in the United States today--are derived from this general system. Includes a list of suggestions for further reading.

18. "The Underground Economy: An Introduction.' May and July 1984 SURVEY. In May, discusses the coverage of the term "underground economy,' a synopsis of measurement methods and results, and a survey of implications. In July, discusses illegal activities in the context of the NIPA's, three sets of NIPA estimates sometimes misunderstood as being measures of the underground economy (including the adjustments for misreporting described in item 20), and the effect on NIPA estimates of possible misreporting in source data due to the underground economy. An appendix in July shows the types of source data used to estimate the income and product components of GNP for a benchmark year. Includes a bibliography.

19. "The U.S. National Income and Product Accounts: Revised Estimates, 1981-83, and First and Second Quarters 1984.' July 1984 SURVEY. Presents a regualr annual revision. Describes methodological changes affecting service components of net exports (different data sources for foreign travel and the separation of remitted and reinvested earnings of unincorporated affiliates of U.S. residents and of foreign residents), quarterly patterns of farm inventories and cash receipts for crops, quarterly pattern of rental income of persons, and deflation of the space rent portions of personal consumption expenditures and housing output (a price index prepared by BEA). In describing the revisions, presents a summary methodology for net interest and personal interest income.

20. "Improved Adjustments for Misreporting of Tax Return Information Used to Estimate the National Income and Product Accounts, 1977.' June 1984 SURVEY. Describes the use of tax return information in the NIPA's and the methodology for preparing improved adjustments for misreporting. The improved adjustments, sometimes referred to as the "underground economy adjustments,' were included in personal consumption expenditures, gross private domestic investment, and several components of charges against GNP. The estimates for 1977, subsequently extended forward and backward to provide a time series, were included in the comprehensive revision released in 1985.

21. State Personal Income: Estimates for 1929-82 and a Statement of Sources and Methods. 1984 (GPO Stock No. 003-010-00125-9, price $9.50). Provides a comprehensive statement of sources and methods used for estimating State personal income, much of which is relevant to the national estimates of income. (A major improvement has since been made in the methodology for nonfarm proprietors' income; see the August 1986 SURVEY). Also presents two case studies of improvements over time--one referring to the definition of other labor income and one referring to the extension of coverage of unemployment insurance data in estimating wage and salary disbursements.

22. "Special Note.--Profits of Financial Corporations.' November 1983 SURVEY. Includes a summary of coverage, concepts, and methodology.

23. "Implicit Price Deflators for Military Construction.' November 1983 SURVEY. Presents some of the results of the project described in item 38.

24. "The U.S. National Income and Product Accounts: 1980-82 and First Quarter 1983.' July 1983 SURVEY. Presents a regular annual revision. Describes methodological changes affecting a services component of net exports (a different data source for ocean freight and port charges), net interest paid by the rest of the world (refinements to the methodology of applying interest rates to the amounts outstanding in asset categories), and deflation of residential investment (a different price index for additions and alterations). In describing the revisions, presents summary methodologies for rental income of persons and rent of houses in personal consumption expenditures, for interest and imputed financial services (see item 7 for an updated methodology), and for proprietors' income.

25. "Net Exports of Goods and Services, 1980-82.' March 1983 SURVEY. Includes a summary methodology for exports and imports of goods and services in current and constant dollars. (See item 3 for an updated, fuller description.)

26. "Special Note.--National Defense Purchases.' November 1982 SURVEY. Includes a summary methodology. (See item 38 for a fuller description.)

27. "National Defense Purchases: Detailed Quarterly Estimates, 1977-82.' November 1982 SURVEY. Introduces quarterly estimates at a more detailed level than previously available, based on the methodology summarized in item 26.

28. "Special Note.--Reinvested Earnings of Incorporated Affiliates in the National Income and Product Accounts.' September 1982 SURVEY. Includes a summary methodology.

29 "The U.S. National Income and Product Accounts: Revised Estimates, 1977-81 and First Quarter 1982.' July 1982 SURVEY. Presents an annual revision (covering 5 years). Describes methodological changes affecting quarterly estimates of gasoline purchases in personal consumption expenditures (a different indicator of volume), the space rent portion of personal consumption expenditures and rental income of persons (introduction of the average contract rent for tenant-occupied units to estimate the average rental value), net interest paid by the rest of the world (introduction of several different interest rates in the estimation of an appropriate interest rate to apply to amounts outstanding in the categories of claims and liabilities reported by U.S. banks), and compensation of State and local government employees (different data on wages and salaries).

30. "Special Note.--The Commodity Credit Corporation in the National Income and Product Accounts.' January 1982 SURVEY. Includes a summary methodology.

31. "Constant-Dollar Manufacturing Inventories.' November 1981 SURVEY. Describes the methodology for manufacturing inventories introduced in the comprehensive revision released in 1980. Introduces the constant-dollar estimates of manufacturing inventories by stage of fabrication made possible by the improved methodology.

32. "Special Note.--Personal Interest Income.' September 1981 SURVEY. Includes a summary methodology.

33. "International Transactions in Measures of the Nation's Production.' May 1981 SURVEY. Introduces the measures of command-basis GNP and terms of trade; discusses the scope and deflation of international flows of factor income, focusing on the addition of reinvested earnings of incorporated foreign affiliates to the factor income flows and on the use of the deflator for net domestic product (both introduced in the comprehensive revision released in 1980); and describes the methodology for deflating other components of exports and imports (see item 3 for an updated, fuller description).

34. "The National Income and Product Accounts of the United States: An Introduction to the Revised Estimates for 1929-80.' December 1980 SURVEY. Presents a comprehensive revision. Describes definitional and classificational changes, most importantly the change that adds reinvested earnings of incorporated foreign affiliates of U.S. direct investors and subtracts the reinvested earnings of incorporated U.S. affiliates of foreign direct investors from the factor income components of net exports. Also describes the elimination of capital gains from earnings of unincorporated foreign affiliates, the reclassification of earnings of unincorporated foreign affiliates to dividends, and the reclassification of hotels and motels to nonresidential structures. Introduces several new aggregates, including gross domestic purchases. Describes methodological changes affecting income components (a number of changes, affecting proprietors' income, corporate profits, and net interest), producers' durable equipment (use of the abbreviated commodity-flow method alone rather than as part of an average), net exports (a new treatment of gold--see item 37), and deflation procedures. The changes in deflation procedures affected personal consumption expenditures (a new price index for hospital expenditures), change in business inventories (see item 31), net exports (including a conceptual change for factor services--see item 33), government purchases (more detail for national defense--see item 38--and for State and local), and the statistical discrepancy. Presents a table indicating new NIPA tables and items.

35. "Federal Fiscal Programs.' February 1980 SURVEY. Includes a discussion, built around reconciliation tables, of the differences between unified budget concepts, as presented in the "Budget of the United States,' and NIPA concepts of receipts, expenditures (including purchases), and national defense purchases. (The tables are updated annually.)

36. "Monthly Estimates of Personal Income, Taxes, and Outlays.' November 1979 SEUVEY. Prepared when the introduction of monthly estimates of four kinds of outlays-- personal taxes and nontax payments, personal consumption expenditures (in current and constant dollars), interest paid to consumers, and personal transfer payments to foreigners--completed the monthly presentation of all components of the personal income and outlay account. Describes the methodology for the annual and monthly estimates (quarterly estimates are the sum of monthly estimates) of personal income and its components; of personal tax and nontax payments, personal outlays, and personal saving; and of the constant-dollar estimates and related series.

37. "Revision of NIPA Estimates.' July 1979 SURVEY. Presents a regular annual revision. Describes a change, mainly affecting net exports, in the treatment of gold.

38. Price Changes of Defense Purchases of the United States. 1979 (NTIS Accession No. PB-80217359). Discusses the results of a major project, started in the mid-1970's by BEA in cooperation with the Department of Defense, to develop price indexes at a detailed level using data provided by the Department of Defense.

39. "The Farm Sector.' November 1978 SURVEY. Includes a methodology for farm income and also for farm output and product.

40. "Key Source Data and Projections for National Income and Product Estimates: Third Quarter 1978.' October 1978 SURVEY. Presents tables showing monthly data sources for current-dollar GNP components and for deflation at the time of the preliminary quarterly estimates. (Although not currently published in the SURVEY, the tables are available each quarter.)

41. "Revisions of NIPA Estimates.' July 1978 SURVEY. Presents a regualr annual revision. Describes methodological changes affecting auto repairs in personal consumption expenditures (a new method based on mileage driven), Federal transfer payments (reclassification of Federal payments to nonprofit institutions for training the unemployed, previously treated as a purchase), contributions for social insurance (a new data source for contributions by State and local governments to pension plans for their employees), and deflation procedures. The changes in deflation procedures affected personal consumption expenditures (for brokerage charges and investment counseling, a new BEA price series; for food, improved weights for the composite price index and new information on imported food) and change in business inventories (for farm inventories, improved weights).

42. "Reconciliation of Quarterly Changes in Measures of Prices Paid by Consumers.' March 1978 SURVEY. Describes three measures: The implicit price deflator for personal consumption expenditures and the chain price index for personal consumption expenditures, both prepared by BEA, and the Consumer Price Index, prepared by the Bureau of Labor Statistics. The reconciliation, which was based on the Consumer Price Index available prior to January 1978, has been discontinued.

43. "Revisions of the NIPA's.' July 1977 SURVEY. Presents a regular annual revision. Describes methodological changes affecting the deflation of fixed investment and government purchases (introduction of a lagging procedure for one of the construction price indexes and revisions in the lags applied to the price indexes for equipment to put them on a delivery basis) and of net exports (application of price indexes at a more detailed level).

44. "Reconciliation of BEA Compensation and BLS Earnings.' May 1977 SURVEY. Describes the two frequently used series on hourly compensation; the reconciliation table continues to be published each quarter in the SURVEY.

45. "Revisions of the NIPA's.' July 1976 SURVEY. Presents a regular annual revision. Describes a methodological change affecting the change in business inventories (use of a more detailed procedure in preparing constant-dollar estimates).

46. "Manufacturing and Trade Inventories and Sales in Constant-Dollars, 1959 to First Quarter 1976.' May 1976 SURVEY. Includes a methodology for constant-dollar stocks of inventories. (Item 31 includes an updated methodology for manufacturing inventories.)

47. "The National Income and Product Accounts of the United States: Revised Estimates, 1929-74.' January 1976 SURVEY, Part I. Presents a comprehensive revision. Describes the definitional and classificational changes, including those made to capital formation and consumption (including the introduction of the capital consumption adjustment --see item 49) and to government transactions (including reclassification of interest paid by the Federal Government to foreigners and introduction of imputed interest receipts and imputed charges for services furnished without payment by financial intermediaries). Describes the methodological changes affecting the motor vehicles components (see item 48), change in business inventories (mainly the introduction of information that permits a better separation of book value inventories according to the accounting method underlying them), producers' durable equipment (valuation of computers leased by the manufacturers to others at cost and inclusion of nuclear reactor fuel elements and control rods), government purchases (a change in procedures for converting information on local governments from a fiscal year basis to a calendar year basis and inclusion of certain receipts as negative State and local purchases), and deflation procedures. The changes in deflation procedures affected personal consumption expenditures (food and financial services), construction (see item 52), producers' durable equipment (see item 50), and government purchases (an approximation of specification pricing for employee services and new detail on purchases from business). Also describes the shift from 1958 to 1972 in the base year for the calculation of constant-dollar estimates and price indexes and the methodology for preparing constant-dollar estimates of nonfactor charges (indirect business tax and nontax liability, subsidies less the current surplus, and business transfer payments) that made possible the presentation of several new constant-dollar aggregates, notably national income. Presents exhibits indicating new NIPA tables and items.

48. "Vehicles: Recent Developments and Treatment in the GNP Accounts,' November 1975 SURVEY. Summarizes the methodology (including improvements introduced in the comprehensive revision released in 1976) used to estimate the auto and truck components of the expenditure components of GNP.

49. "New Estimates of Capital Consumption Allowances in the Benchmark Revision of GNP.' October 1975 SURVEY. Describes the concept and methodology for the adjustment, introduced in the comprehensive revision released in 1976, that converts tax-return-based measures of depreciation to measures that are based on consistent accounting and valued at current replacement cost.

50. "Improved Deflation of Producers' Durable Equipment.' July 1975 SURVEY. Describes the use of the Wholesale Price Index (now Producer Price Indexes) and other price indexes, featuring a timing adjustment to place the indexes on the delivery basis needed for the NIPA's. (As of July 1986, the adjustment was no longer needed for the Producer Price Indexes; see item 7.)

51. Quarterly GNP Estimates Revisited in a Double-Digit Inflationary Economy. BEA Staff Paper No. 25, 1974 (NTIS Accession No. COM 75-10042). Summarizes the sources and methods used to estimate current-dollar quarterly GNP and quarterly national income and to deflate quarterly GNP (tables 2 and 3 in this article are more up to date). Describes the sources and methods used to estimate inventories and the inventory valuation adjustment.

52. "Revised Deflators for New Construction, 1947-73.' August 1974 SURVEY, Part I. Describes the price indexes from government agencies and trade sources used to calculate the deflators for each type of construction activity. (Several changes in methodology were made later.)

53. "U.S. National Income and Product Accounts, 1970-73.' July 1974 SURVEY. Presents a regular annual revision. Introduces two new tables: A reconciliation of national defense outlays in the unified budget and NIPA national defense purchases (see item 35) and a reconciliation of the Commodity Credit Corporation's outlays in the unified budget and its NIPA expenditures (see items 30 and 8).

54. Reliability of the Quarterly National Income and Product Accounts of the United States, 1947-71. BEA Staff Paper No. 23, 1974 (NTIS Accession No. COM-74-11538). Discusses possible sources of error in the quarterly estimates and presents four ways of assessing reliability, giving most weight to measures of revisions. An appendix summarizes data sources used in preparing NIPA estimates (table 2 in this article is more up to date).

Publications for which an NTIS accession number is given can be ordered from the National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161. Publications for which a GPO stock number is given can be ordered from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. See "A User's Guide to BEA Information' in the February 1987 SURVEY for additional information about BEA's publications.

Table: 1.--National Income and Product Account, 1986

Table: 2.--Principal Source Data and Estimating Methods Used in Preparing Current-Dollar Estimates of GNP

Table: Commodity-Flow Method

Table: Retail-Control Method

Table: Perpetual-Inventory Method

Table: 3.--Methodology Used in Preparing Constant-Dollar Estimates of GNP

Table: 4.--Directory to Information about GNP