Integration of U.S. Macroeconomic Accounts: A Progress Report

56 Business Economics • April 2005 Focus on Statistics

F O C U S O N S T A T I S T I C S

Integration of U.S.MacroeconomicAccounts: A ProgressReport

By: Robert P. Parker

Robert P. Parker is Chief Statisticianof the U.S. GovernmentAccountability Office.

The Commission of the EuropeanCommunities et al. (1993), onbehalf of the United Nations,

and the International Monetary Fund(1993) developed international guide-lines for macroeconomic accounts.They recommended the preparationfor each major sector of an integratedset of current accounts (productionand income), accumulation accounts(capital, financial, and other changesin volume), a revaluation account, anda balance sheet. The existing U.S.accounts, parts of which are preparedby the Bureau of Economic Analysis(BEA) and the Federal Reserve Board(FRB), are not complete and are notconsistent because in some cases thetwo agencies use different estimatingmethodologies and different defini-tions. Last year, the two agenciesreported on a major collaborativeeffort to expand the accounts, resolvemethodological differences, andimprove consistency with the interna-tional guidelines. (Teplin, et al., 2004)The feature of their report was a draftset of complete accounts for 1985-2002. In addition, BEA published anarticle in December 2004 on itsefforts to improve consistencybetween the national income andproduct accounts (NIPAs) and theinternational guidelines. (Mead, et al.,2004).

This article reviews the results of

this joint agency effort to providecomplete and integrated U.S.accounts. It discusses the new dataprepared as a result of this effort andthe proposed work plan to eliminatesome of the major differences withinthe U.S. accounts and between theU.S. accounts and the new interna-tional guidelines.

BackgroundInternational guidelines

The international guidelines fornational economic accounting werereleased in 1993 by the UnitedNations (System of NationalAccounts) and the InternationalMonetary Fund (Balance of PaymentsAccounts). The System of NationalAccounts (SNA) provides a compre-

hensive framework for recording allof the stocks and flows in a nation’seconomy, including internationaltransactions. It also covers financialaccounts and balance sheets as wellas input-output tables. The balance ofpayments manual (BPM), which isfully consistent with the SNA, pro-vides similar accounts for interna-tional transactions.

The SNA is designed as an inte-grated system in that it uses consis-tent definitions, classifications, andaccounting conventions for all sec-tors of the economy. As shown inFigure 1, for each sector, the SNAsummarizes the transactions ofgroups of institutions into a sequenceof accounts that flow from one toanother—e.g., production generates

S N A S T R U C T U R E O F I N T E G R A T E D A C C O U N T S

OPENING BALANCE SHEET

• (Value of assets, liabilities, and net worth)

CURRENT ACCOUNT

• Production account (GDP, consumption of fixed capital, compensation of

employees, taxes on production and incomes less subsidies, net operating

surplus)

• Distribution and use of income (income and outlay)

ACCUMULATION ACCOUNT

• Capital account (net fixed investment and net lending/borrowing)

• Financial account (net acquisition of financial assets and net incurrence

of liabilities)

• Other changes in volume account

• Revaluation account (changes in net worth from holding gains/losses)

CLOSING BALANCE SHEET

• Changes in stock positions (changes due to capital formation, net

lending/borrowing, revaluation of assets and liabilities, other changes in

volume)

•Closing position (value of assets, liabilities, and net worth)

F I G U R E 1

Business Economics • April 2005 57Focus on Statistics

income, which is used to financeconsumption and saving. Saving isused for capital formation or foracquiring financial assets and liabil-ities. The accumulation of non-finan-cial and financial assets and liabili-ties and the revaluation of assets andliabilities explain the differencesbetween opening and closing bal-ance sheets. All together, the inte-grated accounts track the sources ofchange in a sector’s net worth. Thefive major SNA sectors are non-financial corporations, financial cor-porations, general government, non-profit institutions serving house-holds, and households. The BPMprovides a full set of integratedaccounts for the rest-of-the worldsector that are fully consistent withthe SNA.

The United States and most othercountries are in various stages of pro-viding a fully integrated set of macro-economic accounts. For example,among the developed countries,Australia, Canada, and the UnitedKingdom prepare integrated accounts,although they do not cover all of theSNA sectors; and there are some dif-ferences in the definitions of sectors.1

In Europe, the European CentralBank is coordinating development ofquarterly financial accounts for theEuro area for the non-financial sec-tor, insurance companies, and pen-sion funds and is working to expandthe integrated system to other sec-tors.

Looking ahead, the United Na-tions Statistical Commission isreviewing the SNA and related guide-lines, such as those in the BPM, for arevision to be released in 2008.Among the proposals currently underconsideration are the following: treatspending on defense weapons systemsas investment in fixed assets, include

employee stock options as compen-sation, recognize a net return to gov-ernment-owned fixed assets, includegovernment-guaranteed loans in thebalance sheet, and treat expendi-tures on research and developmentas capital formation.2

Changing the U.S. accountsIntegrating the U.S. accounts to

facilitate various types of analysishas a long history. For many years,economists have called for increasedconsistency between the differentU.S. accounts and for a complete setof related balance sheets. In theearly 1980s, Richard and NancyRuggles (1982) prepared a set ofaccounts that related the income andproduct flows to balance sheets.3

Commenting on their effort, JamesTobin noted “their experiment …illustrate[d] the well-known problem.It is difficult to reconcile data fromthe different sources, and disturbing-ly large, unexplained discrepanciesremain …. Conceptual integrationneeds to be matched by a concertedeffort to diagnose and remedy theseinconsistencies.” Subsequently, RobertEisner (1986) and Michael Boskin, etal. (1989), among others, wrote on theneed for integration. Although thesecritics recognized that the existingU.S. accounts covered many ele-ments of such a set of integratedaccounts, the missing pieces werecritical for many types of analysis.Moreover, because these accountsare presented in different publica-tions and use different terminologyfor the same concept, combiningthem was difficult, even for the

sophisticated user. The criticsthought that integration of theaccounts would provide a commonterminology and a uniform presenta-tion that highlighted connectionsbetween the activities described inseparate accounts.

Publication of the revised inter-national guidelines in 1993 resultedin calls for the United States toimprove the consistency between theU.S. accounts and these guidelinesin order to facilitate internationalcomparisons. For example, in 1994,the National Academy of Sciencesconvened a panel to discuss integra-tion of the NIPAs and flow of fundsaccounts (FFAs) for the governmentsector. The panel recommended thatBEA “continue to move toward adop-tion of the major features of the inter-national System of National Ac-counts (SNA) accounting frameworkfor the government sector …” andthat the FRB “working closely withBEA, should develop and regularlypublish balance sheets for the gov-ernment sector” (Slater and David,1998). Shortly thereafter, BEA con-ducted a strategic review of itsnational, international, and regionalaccounts and, with substantial inputfrom users, developed a five-yearstrategic plan. This plan not onlyincluded work with the FRB to betterintegrate the two agencies’ accountsbut also included work for BEA tomake its accounts more consistentwith the 1993 international guide-lines (Bureau of Economic Analysis,1995). Users participating in thereview thought that closer coordina-tion between the agencies wouldensure that certain critical ele-ments—such as sector boundaries,alternative data sources, and treat-ment of transactions—would be han-dled in a way that minimized distor-tions to important analytical con-cepts. Also, integration would alignU.S. statistics more closely with those

1For a report on the integration by these coun-tries, see Wilson (2004).

2Detailed information on the process and pro-posals is available on the United Nations website at www.unstats.un.org/unsd/nationalaccounts/snarev1.htm. To provide comments toBEA, contact Brent Moulton at [email protected]. 3Discussion by other experts follows their arti-cle in the same issue of the Survey of CurrentBusiness.

of other nations and allow policymak-ers and researchers to analyze morefully and accurately the interrelation-ships of the nation’s financial andnon-financial activities.

Beginning with the comprehen-sive NIPA revision of 1996, BEA hasbeen implementing changes to boththe NIPAs and international transac-tions accounts (ITAs) to increase con-sistency with the international guide-lines. For example, BEA adoptedchain-type indexes for measuringchanges in real GDP and prices in1996, established a capital accountin the ITAs and redefined fixedinvestment in the NIPAs to includesoftware as investment in 1999, andrecognized implicit services providedby commercial banks to borrowers inthe NIPAs and ITAs in 2003.4 BEAalso changed table designs and rede-fined certain transactions to makethem much more consistent with theSNA.5 The FRB has incorporatedsimilar changes into the FFAs. In themost recent development, BEA andFRB staff presented the paper byTeplin, et al. (2004) last April toreport on progress on a major effort toimprove the integration of the U.S.accounts and their consistency withinternational guidelines. A summaryof this paper is presented below in“Report of Joint Agency Effort.”

Existing U.S. accountsThe existing U.S. macroeconom-

ic accounts are prepared by BEA,who prepares the NIPAs and theITAs, and by the FRB, who preparesthe FFAs.6 The NIPAs provide pro-duction and income accounts for the

United States and data on the accu-mulation and value of tangibleassets. For the production accountand the measurement of GDP, thereare three domestic sectors—busi-ness, general government, andhouseholds and institutions. For theincome and savings accounts, thereare three different sectors—person-al, government, and corporate. In theproduction account, governmententerprises, private unincorporatedbusinesses, and tenant-occupiedhousing are treated as part of thebusiness sector. In the income andsaving account, private unincorpo-rated businesses and tenant-occu-pied housing are classified in thepersonal sector; and governmententerprises are classified in the gov-ernment sector. The ITAs provideU.S. transactions and balances withthe rest of the world and include cur-rent, capital, and financial accountsas well as the net internationalinvestment positions of the UnitedStates.

The FFAs provide capitalaccounts (showing saving and capitalexpenditures), a financial account(showing net acquisition of financialassets and net incurrence of liabili-ties), and balance sheets for selectedsectors—households and nonprofitinstitutions serving households, non-financial corporate business, and non-financial noncorporate business. Forthe other sectors, balance sheet dataare limited to financial assets and lia-bilities. For the financial sector, theFFAs include detail on flows of specif-ic financial instruments, such as mort-gages, corporate bonds, and deposits,and stocks of financial assets and lia-bilities. For the rest of the world sec-tor, the FFAs provide details on finan-cial assets and liabilities.

Although the U.S. accounts arenot prepared following internationalguidelines, the United States hasbeen providing international organi-

zations with macroeconomic accountestimates prepared as closely as pos-sible to international guidelines.These special tabulations enablethese organizations to publish datafor the United States alongside com-parable data from other countries.BEA has been doing this for manyyears, although a lack of source dataprevents it from eliminating all dif-ferences with the SNA. The FRBrecently began providing interna-tional organizations with quarterlyFFAs prepared using definitions andterminology more consistent withinternational guidelines.

Differences between U.S. accountsand international guidelines

Although changes have beenmade to the U.S. accounts to makethem more consistent with interna-tional guidelines, many major differ-ence remain. For example, theNIPAs and FFAs treat weapons sys-tems as investment and not con-sumption as in the SNA; and theFFAs treat household purchases ofdurable goods as investment and notconsumption (Fraumeni and Okubo,2001). There also are differences incoverage primarily due to the lack ofsource data. For example, the SNAtreats illegal activities (such as pros-titution or the cultivation or manu-facture of illegal drugs) the same aslegal activities, but the U.S. accountsdo not cover them. The SNA alsotreats expenditures for livestock andentertainment—literary or artisticoriginals (such as original films,sound recordings, and manu-scripts)—as investment, not con-sumption as in the existing U.S.accounts.

Important differences related tothe definition of the sectors alsoremain. Although the SNA, theNIPAs and FFAs all define their sec-tors in terms of groups of institutions,they differ in the characteristics used


4For a summary of these changes, see Mead, etal. (2004)5Changes in the NIPAs are described byMoulton and Seskin (2003) and byMayerhauser, et al. (2003). 6Detailed information on the accounts pre-pared by these agencies is available atwww.bea.gov for BEA and www.frb.gov forFRB.

to determine these groups. For exam-ple, in the SNA, all institutions thatproduce goods and services for saleat a price that covers the cost of pro-duction and maintain a complete setof books are classified into the non-financial or financial corporationsectors without regard to their owner-ship, legal form of organization, orpurpose. In the U.S. accounts,greater emphasis is placed on owner-ship, legal form, and purpose; conse-quently, corporations organized forprofit, all government agencies, non-profit institutions serving house-holds, and unincorporated privatebusinesses are almost always classi-fied in the same FFA or NIPA sector.(Because the NIPAs have differentsectors for the production accountand the income and outlay account,there are differences in the NIPAaccounts in the treatment of govern-ment agencies and certain unincor-porated businesses.)

Finally, another major differencebetween the SNA and the U.S.accounts relates to statistical dis-crepancies that arise in the U.S.accounts but not in the SNA. In theU.S. accounts, there are several well-know discrepancies–such as the dis-crepancy between GDP as measured(a) as the sum of expenditures and (b)as the sum of incomes in the NIPAsand personal saving in the NIPAsand FFAs. The SNA avoids such dis-crepancies by recommending thatcountries use balancing items andallocate (force) any discrepancieswithin the accounts (Commission ofthe European Communities, et al.,1993, paragraphs 3.64 and 3.65, p.78). Although the elimination of sta-tistical discrepancies in U.S.accounts may improve consistencywith international guidelines, somediscussants at the conference wherethe paper was presented expressedconcerns about such eliminationbecause they used the discrepancies

to assess the reliability of thesemeasures. (Slifman, 2004).

Report of Joint Agency EffortGoals of effort

The authors of the April jointagency paper stated the goals of theeffort as follows:

“Diagnosing and remedyinginconsistencies has been ourgoal. Indeed, preparing thispaper, including the attachedtables, is perhaps the most sig-nificant joint effort on theaccounts since the Tobin com-ment was made. The agencieshave looked closely at sectorboundaries and the nature ofdiscrepancies that arise fromusing different data sources,judgmental adjustments, andestimating techniques. Wehave uncovered many issuesand have solved a few; othersremain for future work”(Teplin, et al., 2004, p 3).

The authors identified many keyinconsistencies between the existingBEA and FRB accounts, as well asbetween these accounts and interna-tional guidelines, and presented afull set of integrated U.S. accountsfor 1985-2002, which are referred toas “Draft SNA-USA.”7

InconsistenciesTeplin, et al. discuss two types of

inconsistencies in the existing U.S.accounts: (a) inconsistencies due tomethodological differences, includ-ing the use of different source dataand the use of different estimatingprocedures with the same sourcedata; and (b) inconsistencies in thedefinitions of the sectors, includingwithin the NIPAs and between the

NIPAs and FFAs. They discuss thedifferences in the source data used toclassify captive finance corporationsand differences in the methodologiesfor using data on state and local gov-ernments. They note differences inthe NIPAs and FFAs classification ofnonprofit institutions serving house-holds and housing as well as the useof two different sets of sectors for theNIPA production and income andoutlay accounts. The paper also dis-cusses other aspects of the U.S.accounts: the FFAs’ other changes inthe volume and the revaluationaccounts lack detail on types of assetsand liabilities, and the revaluationsaccounts do not show separate detailon holding gains or losses. In addi-tion, there are no balance sheets forseveral key sectors; and the NIPAsprovide production accounts byindustry, but not by sector. Moreover,there are two measures of GDP.

Teplin, et al. also discusses theinconsistencies, or differences,between the existing U.S. accountsand the international guidelines.These differences, which are men-tioned above in “Background,” pri-marily relate to the definitions of sec-tors and the lack of a full sequence ofaccounts for all sectors. The SNA,the NIPAs, and the FFAs define theirsectors in terms of groups of institu-tions; but they differ in the charac-teristics used to determine thesegroups. These differences primarilyrelate to institutions, other than cor-porations organized for profit, whichproduce goods and services for saleat a price that covers the cost of pro-duction and maintain a complete setof books. In the SNA, these types ofinstitutions, are called “quasi-corpo-rations” and are classified into eitherthe non-financial or financial corpo-ration sectors without regard to theirownership, legal form of organiza-tion, or purpose. In the U.S.accounts, greater emphasis is placed


7The Draft SNA-USA tables 1 through 9included in the paper show data for 1995through 2002. A complete set of tables for1985-1994 and data for all years is on BEA’swebsite at www.bea.gov.

on ownership, legal form, and pur-pose in determining the classifica-tion by sector.

Government agencies that areclassified as “quasi-corporations,”which are called “government enter-prises” in the U.S. accounts, areclassified for the FFAs and the NIPAincome and outlay account into thegeneral government sector, and forthe NIPA production account into thebusiness sector. All nonprofit institu-tions serving households, includingthose classified in the SNA as “quasi-corporations,” are grouped togetherto form a separate nonprofit institu-tions serving households sector in theU.S. accounts. Unincorporated busi-nesses that are classified in the SNAas “quasi-corporations” are classified

for the FFAs in the nonfinancial non-corporate sector, in the personal sec-tor for the NIPA income and outlayaccount, and in the business sectorfor the NIPA production account.However, the SNA classifies thesebusinesses as “quasi-corporations”only if they maintain completerecords; consequently those unincor-porated businesses without completerecords are classified in the house-hold sector. No such distinction ismade in the U.S. accounts; all unin-corporated businesses are classifiedin the same sector. The Teplin, et al.paper and Mead, et al. (2004) notethat the United States lacks sourcedata to implement these SNA guide-lines. The article also notes that theSNA sector definitions may not pro-

vide important measures, such aspersonal income, that are importantfor U.S. users.

The other significant differencerelates to rental housing owned byindividuals and not businesses. Inthe SNA, this housing is classifiedinto the household sector. In theFFAs, they are classified into thenon-financial noncorporate sector. Inthe NIPAs, they are classified intothe personal sector for the incomeand outlay account and into the busi-ness sector for the productionaccount. The differences for govern-ment enterprises and unincorporatedbusinesses primarily reflect the lim-ited source data available to imple-ment the SNA.

Draft SNA-USAThe Draft SNA-USA accounts

prepared for the joint agency paperprovide close to a full set of SNAaccounts for 1985-2002. As notedabove, Teplin, et al. were not able toprepare sector accounts that werefully consistent with the SNA. Themajor differences are noted inFigure 2.

Despite these differences andlimitations, Draft SNA-USA providesfar more information than the exist-ing U.S. accounts and will hopefullylead to a process of continuingimprovement. In preparing the newaccounts, BEA and FRB haveaccomplished the following:• Expanded the availability of bal-

ance sheets to cover the govern-ment and rest of world sectors

• Provided a continuous timeseries for 1985-2002

• Minimized differences in theexisting U.S. accounts by reclas-sifying personal trusts and cor-porate farms and by improvingtiming adjustments related tofederal government transactions

• Eliminated internal inconsisten-cies such as the difference


M A J O R D I F F E R E N C E S B E T W E E N T H E S N A A N D D R A F T

S N A - U S A S E C T O R S

NON-FINANCIAL CORPORATIONS

SNA-USA includes individually owned rental housing; excludes goverment

agencies and nonprofit institutions serving households that cover costs

with revenue; “large” non financial unincorporated businesses are included in

separate sector

FINANCIAL CORPORATIONS

SNA-USA includes “small” unincorporated businesses and excludes

government agencies and nonprofit institutions serving households that

cover costs with revenue

GENERAL GOVERNMENT

SNA-USA includes government agencies that cover costs with revenue

enterprise

HOUSEHOLDS

SNA-USA shows combined household and nonprofit institutions serving

households sector; excludes individually owned rental housing; “small”

unincorporated businesses are included in separate sector

NONPROFIT INSTITUTIONS SERVING HOUSEHOLDS

SNA-USA not shown separately

REST OF THE WORLD

No major differences

F I G U R E 2

between GDP measured fromexpenditures and incomes

• Presented the new accountsusing a standard format and ter-minology more consistent withthe SNA

At the same time, the newaccounts have several limitations, inaddition to the lack of complete con-sistency with the SNA sectors. Forexample, the statistical discrepan-cies related to the estimates of netlending or net borrowing remainlarge, and the new balance sheets forthe government sectors do notinclude the value of land. Neverthe-less, Draft SNA-USA has providedusers with a set of accounts that morefully reflect international guidelineson presentations, has eliminated

some of the inconsistencies in theexisting U.S. accounts, and is pre-sented in an integrated framework.The authors, in assessing the degreeof consistency in the new estimates,compared the relative size of the dif-ferences in the net lending/net bor-rowing as measured in the new capi-tal and financial accounts in theDraft SNA-USA sector accounts. Inthe capital account, net lending/netborrowing is measured as net savingand capital transfers less net capitalformation, based on the NIPAs andITAs, adjusted to eliminate severalinconsistencies. In the financialaccount, net lending/net borrowing ismeasured as net acquisition of finan-cial assets less net incurrence of lia-bilities, based on the FFAs. The dif-ferences, or the “statistical discrep-

ancy,” primarily reflect differencesin source data, the timing in record-ed flows, and other statistical differ-ences between the U.S. capital andfinancial accounts.

For five of the sectors covered bythe new estimates, the estimates inTable 1 show that there are signifi-cant differences in the patterns andsize of the discrepancies.8 For exam-ple, Teplin, et al. reported that thedifferences in net lending/net bor-rowing for the household sector aresignificant, but the patterns are sim-ilar except for the last several years.


C O M P A R I S O N O F N E T L E N D I N G / N E T B O R R O W I N G I N D R A F T S N A - U S A : S E L E C T E D S E C T O R S

(Billions of dollars)

Sector 1985 1990 1995 1999 2000 2001 2002

Household and nonprofit institutions

serving households:

Capital account 162.8 152.1 67.7 -147.7 -153.6 -209.9 -171.0

Financial account 183.9 313.1 118.5 -155.8 -269.2 -9.7 -280.6

Statistical discrepancy -21.1 -161.0 -50.9 8.1 115.6 -200.2 109.7

Non-financial corporate business:

Capital account -11.5 -25.7 -36.7 -124.3 -205.5 -75.3 45.2

Financial account -167.7 -59.2 35.5 -17.7 -28.2 82.4 14.4

Statistical discrepancy 156.2 33.5 -72.3 -106.6 -177.3 -157.8 30.8

Financial business:

Capital account -17.5 2.3 23.2 15.6 14.4 89.3 124.5

Financial account 2.2 -6.1 68.7 52.2 -136.7 -19.1 -10.1

Statistical discrepancy -15.3 8.4 -45.5 -36.5 151.1 108.4 134.6

Federal government:

Capital account -213.3 -208.2 -199.3 101.3 189.3 45.4 -258.7

Financial account -246.2 -193.9 180.4 105.4 198.0 -2.3 -269.3

Statistical discrepancy 33.0 14.3 -18.8 -4.1 -8.7 47.7 10.6

State and local government:

Capital account 1.6 -37.7 -33.0 -22.3 -30.3 -61.0 -86.9

Financial account -9.2 -23.5 0 9.9 -7.1 -39.5 -81.7

Statistical discrepancy 10.8 -14.2 -33.1 -32.1 -23.2 -21.5 -5.2

Source: Teplin, et al. 2004 and BEA website www.bea.gov.

T A B L E 1

8The paper shows detailed data for these sec-tors, as well as for non-financial noncorporatebusiness and rest of the world sectors, for1995-2002 and charts for net lending or netborrowing for most sectors beginning with1985. Data for all sectors for all years are avail-able on BEA’s website.

The two measures showed about thesame net borrowing for 1999, and thefinancial account measure showedmore net borrowing in 2000 and2002. In 2001, the capital accountmeasure showed much larger netborrowing. The authors noted thatthe residual calculation of some cat-egories in the current account leaveconsiderable room for speculation onthe sources of the difference.9 This isalso true for the residual calculationof most asset categories and someliability categories in the financialand balance sheet accounts

There are larger relative differ-ences in net lending/net borrowingfor the non-financial corporate busi-ness sector. In most years, bothmeasures were consistent as towhether the sector was a net lenderor net borrower. Since 1990, therehave been four years where themeasures differed in this regard. In2000, the capital account measureshowed significantly more net bor-rowing. Teplin, et al. noted theboundary between the non-financialand financial business sectors as alikely major cause for the discrepan-cy, in part reflecting differences insource data for the NIPAs and theFFAs. For the current accounts, theNIPAs rely heavily on tax returndata. For the financial account, how-ever, the FFA uses tax return data aswell as data from surveys and regula-tory information. In the NIPAs, cor-porations that file consolidatedreturns that include both non-finan-cial and financial subsidiaries arerecorded either in the financial ornon-financial business sectors,depending on the predominant busi-ness. In the financial accounts,adjustments are made with the sup-plementary information to split thedata from these returns into separate

financial and non-financial corpora-tions. The issue is particularlyimportant for non-financial firmswith captive finance companies.Evidence of this issue being a majorsource of the difference is that thenet lending/net borrowing estimatesof the two sectors combined showmore consistency. Draft SNA-USAdid not include any adjustment forthis boundary issue.

In addition to this boundaryproblem, Teplin, et al. reported thatpart of the difference between netlending/net borrowing measures inthe non-financial corporate businesssector likely reflects the recording of“miscellaneous” financial assets inthe financial accounts. At present,the FFAs include changes to good-will and other intangible assets inflows of miscellaneous financialassets. Such flows are large and pos-itive during periods of heavy mergeractivity and large and negative dur-ing periods of economic weakness.The impact of changes in such assetsis not reflected in the Draft SNA-USA current and capital accounts.The adoption of the internationalguidelines would result in the alloca-tion of a portion of the changes in thevalue of such assets to the otherchanges in volume account.Although the reclassification of theseassets into revaluations and/or otherchanges in volume is likely to have asizable impact on the statistical dis-crepancy, changes were not made inDraft SNA-USA for these accountingissues.

For government sectors, DraftSNA-USA shows mixed results. Forthe federal government sector, netlending/net borrowing estimates areclose for all years, reflecting a gener-ally high quality of available sourcedata. These estimates also reflectchanges incorporated into Draft SNA-USA to improve the use of these data,especially when special timing

adjustments, such as for the paymentof unemployment insurance benefits,are needed for the current account.However, net lending/net borrowingestimates for the state and local gov-ernment sector show significant dis-crepancies. For this sector, the sourcedata used for the U.S. accounts tendto be less timely and of lower quality.The authors concluded that the coor-dination of estimating proceduresusing the available state and localgovernment data could reduce thesize of these discrepancies.

Next StepsThe joint effort of BEA and FRB

staffs has produced a first set of inte-grated sector accounts for the UnitedStates following, to the extent possi-ble, international guidelines alongwith the recommendations presentedin April 2004 by Teplin, et al. Theauthors reported that additionaleffort would be needed to producethe estimates on a continuing basisor extend them for the period before1985. They also reported that an evengreater effort would be needed toimplement improvements to the U.S.accounts that would eliminate incon-sistencies within these accounts,expand them to provide a full set ofaccounts, or improve their consisten-cy with the international guidelines.

With regard to the elimination ofinconsistencies within the U.S.accounts, Teplin, et al. noted theproblems related to the differentmethodologies for the treatment ofcaptive finance corporations andpreparation of estimates on state andlocal governments, the inconsistencyof the definition of sectors within theNIPAs, and the reclassification ofpersonal bank trust and miscella-neous financial assets in the FFAs.To expand the coverage of the U.S.accounts, the authors discussed theneed to develop further the otherchanges in volume and revaluation


9For a discussion of these differences, seeWilson, et al. (1989).

accounts through a review of theaccounting techniques used insource data and whether those tech-niques align with what is needed forthe accounts. To provide completebalance sheets, they noted the needto develop improved source data forreal estate values, especially for thegovernment sector, and improvedmethodologies for separating themarket value of land and structures.Finally, improving consistency withinternational guidelines, especiallythe consistency of the definitions ofthe sectors, were linked primarily tothe availability of improved sourcedata.

Teplin, et al. reported that closercoordination between the agencieswould be needed to ensure that cer-tain critical elements—such as sec-tor boundaries, alternative datasources, and treatment of transac-tions—would reduce the inconsis-tencies in the existing accounts.BEA’s Strategic Plan outlines a num-ber of research and developmentactivities related to the developmentof integrated accounts. Work isalready underway to examine sectordefinitions, with particular consider-ation given to the development ofimproved source data on governmentbusiness enterprises. Other researchis also being conducted on improvingthe consistency of source databetween the NIPAs and FFAs, forexample, to improve consistency ofNIPA interest flow estimates withFFA estimates of interest-bearingassets and liabilities. In addition tothe joint staff work noted above, theauthors noted that efforts at the FRBare directed toward improving inte-gration in the capital accounts, espe-cially disaggregating the combinedfinancial sectors. In conjunction withthat work, there will be an effort to

develop detailed information for theother changes in volume accounts.The authors proposed future work toeliminate additional types of incon-sistencies, to enhance the presenta-tion of the existing accounts andDraft SNA-USA, and to producenational balance sheets on a regularbasis. However, progress will dependon appropriate allocation of staff bythe two agencies to the joint effort. ■

R E F E R E N C E SBoskin, Michael J., Marc S. Robinson,

and Alan M. Huber. 1989. “GovernmentSaving, Capital Formation, and Wealth In theUnited States, 1947-85.” The Measurement ofSaving, Investment, and Wealth Chicago.

Bureau of Economic Analysis. 1995.“Mid-Decade Strategic Review of BEA’sEconomic Accounts: An Update.” Survey ofCurrent Business (April) 48-56.

Commission of the European Communi-ties-Eurostat, International Monetary Fund,Organization for Economic Co-operation andDevelopment, United Nations, and WorldBank. 1993. System Of National Accounts1993 Prepared under the auspices of theInter-Secretariat Working Group on NationalAccounts.

Eisner, Robert. 1986. How Real is theFederal Deficit? New York: The Free Press.

Fraumeni, Barbara M. and SumiyeOkubo. 2001. “Alternative Treatments ofConsumer Durables in the NationalAccounts.” Paper prepared for the BEAAdvisory Committee Meeting. May 11.

International Monetary Fund. 1993.Balance of Payments Manual (fifth edition).Washington, D.C.

Mead, Charles Ian, Karin E. Moses, andBrent R. Moulton. 2004. “The NIPAs and theSystem of National Accounts.” Survey ofCurrent Business. (December) 17-32.

Mayerhauser, Nicole, Shelly Smith, andDavid F. Sullivan. 2003. “Preview of the 2003Comprehensive Revision of the NationalIncome and Product Accounts: New andRedesigned Tables.” Survey of CurrentBusiness (August) 7-31.

Moulton, Brent R. and Eugene P. Seskin.2003. “Preview of the 2003 ComprehensiveRevision of the National Income and ProductAccounts: Changes in Definitions and

Classifications.” Survey of Current Business(June) 17-34.

Ruggles, Richard and Nancy D. Ruggles.1982, “Integrated Economic Accounts for theUnited States, 1947-80.” Survey of CurrentBusiness (May) 1-53.

Slater, Courtenay and Martin David Eds.1998. Measuring the Government Sector of theU.S. Economic Accounts. Washington D.C. 9.

Slifman, Larry. 2004. “Panel Discussion”in Dale W. Jorgenson, J. Steven Landefeld, andWilliam D. Nordhaus, eds. A New Architecturefor the U.S. National Accounts. PapersPresented at a NBER Conference held April16-17, 2004. University of Chicago Press(forthcoming).

Teplin, Albert M., Rochelle Antoniewicz,Susan Hume McIntosh, Charles Ian Mead,Karen Moses, Brent Moulton, MichaelPalumbo, and Genevieve Solomon. 2004.“Integrated Macroeconomic Accounts for theUnited States: Draft SNA-USA.” in Dale W.Jorgenson, J. Steven Landefeld and William D.Nordhaus, eds. A New Architecture for the U.S.National Accounts. Presented at a NBERConference held April 16-17, 2004.University of Chicago Press (forthcoming).Available on BEA’s website at www.bea.gov.

Wilson, John F.; James L. Freund;Frederick O. Yohn, Jr.; and Walter Lederer.1989. “Measuring Household Saving: RecentExperience from the Flow-of-FundsPerspective.” In The Measurement of Saving,Investment, and Wealth. Robert E. Lipsey andHelen Stone Tice, eds. National Bureau ofEconomic Research, Studies in Income andWealth. 52:101-152.

Wilson, Karen. 2004. “The Architectureof the System of National Accounts: A Three-way International Comparison Canada,Australia, and the United Kingdom” in DaleW. Jorgenson, J. Steven Landefeld, andWilliam D. Nordhaus, eds. A New Architecturefor the U.S. National Accounts. Presented at aNBER Conference held April 16-17, 2004.University of Chicago Press (forthcoming).


Documents

Integration of U.S. Macroeconomic Accounts: A Progress Report