NATIONAL ACCOUNTS: SOURCES AND USES

Use of the national accounts, is a very broad topic covering a wide range of issues concerning almost all economic statistics and the various ways in which the national accounts can be used by the property industry.

The national accounts provide a systematic summary of national economic
activity and have been developed to facilitate the practical application of economic theory. They play a critical role in the management of the economy through their use in policy formulation, economic decision making and the evaluation of economic outcomes for the property industry.

The Australian national accounts are compiled from a wide range of data
sources. Detailed information about the sources and methods used to compile the Australian national accounts is contained in the ABS publication Australian National Accounts: Concepts, Sources and Methods (5216.0).

DATA SOURCES OF THE AUSTRALIAN NATIONAL ACCOUNTS

The relative size of particular aggregates, expressed as a percentage of GDP, is a means of highlighting which data sources have the greatest potential impact on the national accounts. All figures quoted relate to annual current price series for the financial year 1993-94 as published in the 1993-94 edition of Australlan National Accounts: National Income, Expenditure and Product (ABS Cat. No. 5204.0).

The income side of the accounts is broadly comprised of three aggregates:


By far the largest of these, comprising 49.2% of GDP, is wages, salaries and supplements.

Wages and salaries are mainly derived from two ABS surveys, the Survey of
Employment and Earnings (SEE) and the Labour Force Bu ey (LFB). Due to
unavoidable delays in updating new businesses on the ABS Business Register, SEE understates total employment. Consequently, estimates for wages and salaries include:


Net indirect taxes amount to 11.9% of GDP. Data are collected by the Public
Finance Section of the ABS using administrative sources (such as government accounting records and Commonwealth. and State Budget Papers).

Estimates of GOB are compiled separately for the foIlowing categories:


Both private corporate trading enterprises (14.6% of GDP) an unincorporated trading enterprises (10.7% of GDP) employ ATO taxation statistics a their principal data source. A substantial number of adjustments are applied to the raw ATO data in order to bring them onto the conceptual basis required for the national accounts.

GOB of dwellings owned by persons comprised 8.4% GDP in 1993-94.
Although a number of data sources are used to compile estimates for this item, the most important of these is the ABS Census of Population and Housing. The population census is used not only to derive estimates of actual rents pa but also as the basis for the imputed rent attributed to owner occupied dwellings.

Operating expenses relating to dwellings owned by persons are derived from a variety of sources, for example the ABS Household Expenditure Surveys (repairs and maintenance ) and the Office of the Insurance and Superannuation Commissioner (dwelling insurance). Operating expenses are deducted from gross rent in the process of deriving GOB for dwellings
owned by persons.

GOB of public trading enterprises (4.6 % of GDP) and neral government (1.8% of GDP) are relatively minor components of total GOB. GOB public trading enterprises is derived mainly from financial statements attach to the annual reports of PTEs and from Auditors'-General reports. General government GOB estimates are equivalent to general government consumption of fixed capital depreciation at current replacement cost) and are derived using a perpetual inventory model.

Financial enterprises GOB (1.5% of GDP) is built up from a range of data
sources including information from ABS financial enterprise surnveys, the Australian Financial Institutions Commission, the Reserve Bank Statistics -Bulletin, the Insurance and Superannuation Commissioner's statistics and Reports of the State Auditors' General.

The imputed bank service charge (- 2.7% of GDP) is simply that intermediate
consumption relating to financial services for which no explicit charge has been levied by financial institutions. Because of the difficulty involved in allocating this item to individual industries and institutional sectors, it is presently simply allocated to the nominal industry (ie, a "dummy" industry).

SNA93 requires that the imputed bank service charge be allocated to industries and sectors. Data sources are the same as those used for estimating financial enterprises GOS.

On the expenditure side of the domestic production account, GOP is calculated as the sum of:


Final consumption expenditure is broken down into private tin consumption
expenditure (PECE) and government final consumption expenditure.

PFCE equivalent to 61.7% of GDP) comprises a wide range of goods and services purchased by households. A wide array of data sources, both periodic and regular, are used to compile estimates for this aggregate.

However, by far the most important data sources are the periodic retail census (to provide a benchmark for mo of the goods components of PECE) and the monthly Retail Survey (used to interpolate between, and extrapolate from, the benchmark estimates). Other data sources include the Census of
Population and Housing, the Household Expenditure Survey and the Survey of Motor Vehicle Usage. Government final consumption expenditure (equivalent to 18.2% of GDP) is derived from administrative sources such as Commonwealth and State budget papers, Department of Finance ledgers and a joint ABS/Commonwealth Grants Commission annual return from local government bodies.

Private gross fixed capital expenditure (private -GFCE) comprises four
components: dwellings, nondwelling construction, equipment, nd real estate transfer expenses. Private GFCE represents 15.9% of GDP. Private CE on equipment (7.0% of GDP) is mainly derived from ATO taxation statistics, though data from the ABS's quarterly survey of new capital expenditure are used to extrapolate from the latest ATO benchmark.

The ABS Engineering Construction survey and Building Activity Survey provide the main source data for estimates of GFCE nondwelling
construction. Private GFCE on dwelling construction is calculated using information from the ABS quarterly Survey of Building Activity. Taxation statistics, the periodic ABS survey of real estate agents and lawyers and information from the State Stamp Duties Offices provide the main sources of data used in deriving GFC of real estate transfer expenses.

Public GFCE is shown separately for public enterprises and general
government, with each contributing 2.1% to GDP in 1993-94. Public GFCE is estimated from the same sources used to compile estimates of government final consumption expenditure.

Increase in stocks represented only 0.2% of GDP in 1993-94, though the volatility of this item makes it a much more important influence on movements in GOP than this proportion would otherwise suggest. The principal sources of data are: the ABS quarterly Survey of Stocks Owned by Private Enterprises; information provided via returns of the commodity marketing authorities; and estimates ken from the ABS Government Finance System.

Exports and imports of goods and services amounted to 9.3 per cent and
(minus) 19.5 per cent of GDP respectively in 1993-94. Both of these items are
produced through the ABS Balance of Payments System. Data, are derived principally via foreign trade statistics from records maintained by the Australian Customs Service (ACS), though a substantial number of adjustments are made to account for factors relating to timing, coverage and valuation.

In summary, taxation statistics are used principally in relation to GOS of
incorporated enterprises and non-farm unincorporated enterprises and also for private GECE on equipment.


0. NATIONAL ACCOUNTS USE OF AGGREGATE A OlNOOME TAX DATA
FOR BUSINESSES

18. The Nationai Accounts use
Branch of the ABS
e tax data Supplied by the -
ATO for Corn anies, Partnership8 and Trusts, and lndividu Is to provide annual
benchmarks or ustralia for the following aggregates:

* gross operating surplus of private corporate trading - terprises (in total and by
industry);

- gross operating surplus of non4arm Uninqorporated t - ding enterprises (in total
and by industry);

* private gross fixed capital expenditure on equipmdnt n total, by industry and by
institutional sector);

- corporate income tax payable (separately for corpora - trading enterprises and
finanUal enterprises);

- dividends paid by cQrporate trading enterprises and fi ncial enterprises;

- dividends received by corporate trading enterprises a households;

* interest receipts and Payments by industry for corpora trading enterprises and
unincorporated enterprises;

* lease payments in total and by industry (ideally this sh uld only be those relating - --
to finance leases) useq in calculating GOS for corpora trading enterprises and
unincorporated enterprises;

½ * business income for certain "professional" industries, 5 ch as real estate agents,
lawyers and medical practitioners, is used to estimate r extrapolate
- benchmarks from other sources in respect of) some co ponents of private final
consumption expenditure and real estate transfer expe ses.

Fringe benefits tax daja are used to estimate the value of frin - benefits provided to
employees (included in wages and salaries).

19.
ATO income tax data for Individuals and Partnerships a d Trusts, by State and
Territory, are also used in compiling State and Territory estim 5 for gross operating
surplus of Unincorporated enterprises, and various component of household income
and income tax paid which are published in the annual publica n Auskallan National
Accounts: State Accounts (ABS Catalogue No. 5220.0).

20. Data from Company income tax returns. For the cor rate sector the ABS
has generally had access to unit record data although, until the hanges to the Tax Act
in 1986, companies were not identified by name. Unit record d ta in respect of
corupanies are requir9d for many purposes including:

* to distinguish data relating to public trading=enterprise5 ( ese need to be
excluded as data from Public Finance Section are used i preference to ATO
data for these enterprises);

* to distinguish data relating to financial enterprises since ta from other sources
are more suitable for national accounts compilation;



- to enable adjustments to capital expenditure on equi ment to fully account for
finance leasing (which is not recorded by the lessee ATO data and not always
¼ by the lessor either);

- balance date adjustment;

- to estimate for significant missing returns;

- to adjust for large capital losses and/or gains for part ular enterprises.

21. Data from Individuals and Partnerships and Trus income tax returns. For non-farm unincorporated enterprises, aggregate data only h ve been used (Partnerships and Trusts and Individuals), although special bulations are required to obtain separate information for partnerships and trusts in w - h the partners and unit holders are predominantly companies. The data in respect such partnerships and - trusts must be excluded from estimates for the unincorporat - enterprises sector as they will be already included in respect of corporate trading 1nterprises or financial enterpnses.

22. There are a number of issues relating to the use of in Qme tax data in compiling national accounts aggregates, namely;

(a) The quality of industry coding and its consistency wit ABS industry coding are obvious concerns to the National Accounts Branch. cent moves by ATO to adopt ANZSIG industry coding are welcomed.

(b) Adjustments are made to income tax data to allow for- - nderstatement of income (or overstatement of expenses) by business taxpayer - when compiling national accounts estimates.

Ml- - (c) The national accounts estimates for the farm sector a based mainly on ABS data (principally from the annual Agricultural Census wid the Agricultural Finance Survey) rather than income tax data. This avoids ha g to make explicit allowances for understatement of income for tax purp sos for this sector of the economy. A further advantage of this approach is th it ensures consistency between the current and constant price estimates for a farm sector.

(d) - The inclusion of capital gains/losses in the profits dat-available from income taxreturns from companies is a significant problem for national accounts compilailon.

(e) For purchases and disposals of depreciable assets th National Accounts Branch only wants capital equipment since data from ABS su eys are used to estimate capital expenditure on dwellings and nondwelling co truction.

(f) The treatment of finance leasing has been a significa complication because lessees have not been required (in fact, not allowed)- report finance lease payments as capital expenditure while various arrangments have applied for lessors. In some cases lessors have also not been r- uired to report capital expenditure on assets which they have purchased as e lessor in a finance lease arrangement.

(g) In various areas, tax data are potentially a useful sou a of data (in each case in total and by industry), eg, wages and salaries; emplo - r contributions to superannuation; bad debts; and changes in inventorie.

(h) The compilation of separate estimates for the non-pro t institutions sector is a requirement of SNA93. Hence there is considerable i erest in being able to enable adjustments to capital expenditure on equipm t to fully account for
finance leasing (which is not recorded by the lessee in 0 data and not always by the lessor either); balance date adjustment; to estimate for significant missing returns; to adjust for large capital losses and/or gains for particu r enterprises.

21. Data from Individuals and partnerships and Trusts come tax returns. For nonftarm unincorporated enterprises. aggregate data only ha -been used (Partnerships and Trusts and Individuals), although special ta lations are required to obtain separate information for partnerships and trusts in whic the partners and unit -holders are predominantly companies. The data in respect of - uch partnerships and - trusts must be excluded from estimates for the unincorporate nterprises sector as they will be already included in respect of corporate trading e erprises or financial enterprises.

22. There are a number of issues relating to the use of inc e tax data in compiling national accounts aggregates. namely:

(a) The quality of industry coding and its codsistency with BS industry coding are obvious concerns to the National Accounts Branch. Recent moves by ATO to adopt ANZSlG industry coding are welcomed.

(b) Adjustments are made to income tax data to allow for nderstatempnt of income (or overstatement of expenses) by business taxpayer hen compiling national accounts estimates.

(c) The national accounts estimates for the farm sector a- based mainly on ABS data (principally from the annual Agricultural Census- d the Agricultural Finance survey) rather than income tax data. This avoids hav - g to make explicit allowances for understatement of income for tax purp~ sos for this sector of the economy. A further advantage of this approach is th it ensures consistency between~the current and constant price estimates for e farm sector.

(d) - The inclusion of capital gains/losses in the profits dat available from income tax returns from companies is a significant problem for national accounts compilation.

(e) For purchases and disp£sals of depreciable assets t National Accounts Branch
only wants capital equipment since data from ABS 5 veys are used to estimate
capital expenditure on dwellings and non-dwelling co struction.

(f) The treatment of finance leasing has been a signific t complication because
lessees have not been required (in fact, not allowed)o report finance lease
payments as capital expenditure while various arran ments have applied for
lessors. In some cases lessors have also not been quired to report capital
expenditure on assets which they have purchased athe lessor in a finance
lease arrangement

(g) In various areas, tax data are potentially a useful so ce of data (in each case in
-total and by industry), eg, wages and salaries; empl er contributions to
-superannuation; bad debts; and changes in invento - 5=

-(h) The compilation of separate estimates for the non-p fit institutions sector is a
requirement of 3NA93. Hence there is considerabl interest in being able to



identity non-profit institutions in ATO income tax statis os and in ATO coverage
sources which could be used for updating the ABS bu iness register.

-(i) In the past. incomplete processing of income tax retur s, particularly for non-
taxable companies and partnerships and trusts, has b en a significant problem.

23. Several years ago the ABS embarked on a strategy t both improve the source
data available for the national accounts and to provide more omprehensive annual
industry statistics for the whole economy. This strategy invo ed the development of an
annual economy wide collection - the Economic Activity Su {y (LAS). LAS-has two
elements - a completely enumerated large business sector a d a sample of smaller
businesses, A fundamental element of the strategy to impro the national accounts
involves using LAS data for large business in preference to t xation data. The LAS
data are considered to be preferable for the following reason

- (a) Income tax data are only available in respect 0 legal entities, whereas the
LAS data are available for management units ( roducing units delineated
-with the aim of maximising industry homogenei at the 2-digit level of -
ANZSlG for which activity should be largely co ined to a single industry).
This distinction is important for large businesse which typically have
operations spanning a number of industries an a complex structurq of
legal entities.

(b) The definitions of items on the LAS form can b specified to suit national
accounts compilation as far as this is practicab . Although significant
progress has been made towards ensuring tha the statistical data items
on income tax forms are as close as possible t the corresponding items
on the EAS form complete uniformity has not y 1 been achieved. Hence
the LAS provides a better basis for collecting i ormation according to
national accounts requirements.

-\(c) Data from LAS should be more timely than tho - from ATy sources.
9
(d) Data from LAS can be subjected to more rigou us editing than applied by
the ATO and, where necessary, unusual respo ses to LAS questions
could be queried with the respondent.

* 24. For medium and small businesses the arguments in f our of ABS collected data
-are not so strong. In fact, studies undertaken indicate that d-ta reported by these
businesses in ABS surveys align reasonably well with data r orted to the ATO. In
- addition, minimising reporting load for these businesses pro des a compelling
-argument in favour of~using ATO data if practicable.

D. THE NATIONAL ACCOUNTS AND THE UNDEROR UND ECONOMY

The National Accounts Branch is keen to quantify the understatement of income (or overstatement of expenses) in taxation statistics. Howe r, this item represents just
one component of the so-called "hidden" or underground ec omy. To put
understatement of income into its national accounts context, t is necessary to outline
what is meant by the hidden economy. Broadly speaking, t hidden economy
represents all of those production activities that should be in uded in GOP, but which
may be missed because one or more of the parties involved - -ries to hide them from the relevant public authority. More specifically, the hidden econ my could be thought of as comprising the following activities:

--(a) economic activities that are legal in their own ht but which are
deliberately concealed from the authorities so 5 to avoid paying tax or government charges;

(b) production of illegal goods and services; and
(c) employee theft (which should be- recorded as wage and salaries rather than as intermediate consumption).

A failure to cover these activities will result in measure lOP falling short of-
the ideal concept of GOP, that is, as it is defined in 8NA93. Suc - a fdilure will lead to national accounts aggregates which understate the size and pos ibly the growth rate of the economy, which contain biased saving ratios, and which pre nt a misleading picture of the structure of the economy. In terms of the three cat ~gories aboo'e, income
tax statistics will not record the activiW in relation to (a). They wi probably not record a
great deal of income relating to item (b) but they will probably tre t most employee theft
(item (c)) as current business expenses rather than as wages a salaries.

There are a number of approaches to measuring undecla d legal activity (item
(a) above). However, the most common method used world wid is to compare estimates of GOP derived from independent sources, that is, to mpare GDP(l), GDP(E) and, if produced, GlOP(P). It is usually assumed that th - income flow
generated by an-unrecorded (but legal) transaction will be conc led by the recipient
but the expenditure will usually be reported by the purchaser. nsequently, some or
all of the excess of GDP(E) over GDP(l) can be attributed to unr corded legal
transactions. This approach has been used in Australia in conj ction with more direct
estimates of understatement (see paragraph 34 below).

Another approach is to use statistics from income tax au ts. If the audit system
is based on a random sample of taxpayers the under-reporting vealed by the audit
can be applied to the national population of taxpayers. Audits sed on non-random
samples are less useful for this purpose since they are likely to rovide a biased
estimate of the level of understatement. For example, audits th I target a particular
industry or group of taxpayers who are identified as likely to be nderstating their
income are not random sample audits.

29. There have been numerous attempts to determine the e ent of the hidden
economy in various countries by examining the nature of the u of currency within the
economy. At the most basic level this approach involves consi ring the ratio of the
volume of currency in circulation, 0, to demand deposits, 0. A increasing 0/10 ratio is
argued to reflect a greater relative use of cash as the hidden e nomy grows. Of
course, such an approach is sensitive to the base year chosen nd cash holdings may
increase relative to money supply for a number of reasons, not ecessarily connected
with the hidden economy. (Works on this approach include Gu ann (1977) in the UAA
and Tucker (1982) in Australia.) Although a variety of more so histicated econometric
models has been constructed using this particular approach to easure the hidden
economy, such models typically rest heavily on assumptions t t are difficult or
impossible to test (examples of such work include Hepburn (1 2) in Australia and Tanzi (1982), lsachsen et al (1982), and Bhattacharyya (1990) - verseas).

30. A further option is one best described as the "expenditu - approach". it was
used in a paper by N.W.F. Fisher of the Bureau of Labour Mar ;et Research (1983). -
Fisher examined only the expenditure side of the national acc nts. He assumed that
unrecorded transactions in the public sector are likely to be ne ligible in a country like Australia. This allowed him to concentrate on private final con umption expenditure,
private gross fixed capital expenditure and increase in stocks old by the private sector
as potential areas for the "hidden" economy to be exposed. F r each of these three
items (but particularly private final consumption expenditure a d private gross fixed
capital expenditure) the aggregates were dissected as far as ssible. The approach
then involved building up an intimate knowledge of each of th categories for which
esfimates were produced so that judgements could be made out the likely extent of
understatement for each category. While such judgements a - necessarily subjective,
they are likely to be more accurate when applied to individual ems of private final consumption expenditure and capital expenditure than wh n applied at the broad
- - economy level. An advantage of this method is that estim tes for all components of the
hidden economy (ie; items (a) to (c) above) can be jointly stimated.

In relation to item (b) above, SNA93 specifically recommends that production of illegal goods and services be included within the producti - boundary, that is, they
should form pad of GOP. Intuitively, this would include pr uction relating to illegal
drngs, illegal prostitution and illegal gambling. A lack of 5 rce data prevents most
countries from explicitly including illegal production in thei - ational accounts, Those
countries which have tried to measure illegal production u ally employ somewhat
dubious methodologies in attempting to measure a subset] - f total illegal production. For
example, estimates relating to the illegal importation- and bsequent domestic
distribution of cigarettes have been included in the Italian tional accounts, Imports,
wholesale and retail trade and private final consumption e - enditure in the Italian
national accounts have all been adjusted to include the e -cts of the illegal cigarette
trade. However, these estimates have been based mainl n records of seizures of
contraband cigarettes by customs authorities - records th - would intuitively not provide
the means to accdrdtely measure the extent of illegal distr - ution.

Employee theft includes direct theft of materials an supplies by employees from the place of work and office fraud, for example, falsificatio of expense accounts and
the use of employers' property for private purposes. The fect of these activities is to -
understate GOP because goods and services stolen by e - loyees are mistakenly
treated in the employers' accounts (and therefore the nati ial accounts) as intermediate
consumption rather than as wages and salaries. There ar no reliable national statistics
on the extent of employee theft.

UNDERSTATEMiENT OF INCOME IN TAXATION TATISTICS

This section narrows the focus to deal specifically th the understatement of
income (and overstatement of expenses) in ATO taxation atistics. National accounts
estimates of gross operating surplus for the private sector of the economy depend
mainly on income tax data from the ATO (as outlined in P - 5 8 and 0 above).

Consequently any understatement of income in the aggre te income tax statistics will

flow on to the national accounts unless the national accou tants make explicit

adjustments for such understatement. Understatement of- come in the data used to

compile gross operating surplus estimates for private corp rate trading enterprises and

non-farm unincorporated enterprises in the national accou ts can arise from any of the following three possibilities;


(a) understatement of income or overstatement f expenses in respect of

- taxpayers that are included in the data provi d to the ABS; -

(b) omission of information for some taxpayers cause the data are not

included in the tabulations/files supplied to th ABS (eg, data reported to

the ATO but not included in statistical extrac- ns, or late returns not

lodged by the time the statistics are extracte; and

(c) omission of information for some businesses ecause they do not comply

with the requirement to lodge taxation return


34. The ABS has applied an upward adjustment factor the estimates of nonftarm

-unincorporated enterprise income and company income si ce the 1g84-85 issue of the

annual accounts (5204.0). The current system used to adj St for the understatement of

business income in taxation statistics evolved from initial e timates of the revenue

foregone through understatement of business income in t returns which were

produced by the ATO and included in the June 1085 Draft hite Paper "Reform of the Australian Tax System" - business non-compliance. It would be enormously beneficial fo - the ABS to nave access to such data as a means of updating the estimates of uh erstatement of business income in the taxation statistics.


USE OF NATIONAL ACCOUNTS DATA IN MEASURIN NON~COMPLIANCE


38 For internal management purposes the ATO would need 0 regularly estimate the

extent of non-compliance in the business sector in relation to in ome tax and other

taxes such as sales tax and fringe benefits tax- Similarly, esti tes would be required of the extent of non-compliance by individuals in relation to inc e tax. Obviously the

ABS would be interested in those estimates of understatement f business income

relating to both companies and non-farm unincorporated busin ses. If ATO estimates

of business non-compliance are directly or indirectly derived us g national accounts

estimates for incorporated and unincorporated enterprises inco e, then there could be

a problem. Since the ABS derives its estimates of incorporat and unincorporated

enterprises income from taxation statistics and since these esti ates are adjusted upwards to counter inherent understatement of income (on the asis of factors that are

either dated, or at least partially subjective) any attempt by AT - to use national

accounts data to estimate non-compliance in the business sec r will necessanly involve a degree of tirculan'V It is therefore important that due allow - - ce is made for the way

in which the national accounts estimates of incorporated and u incorporated enterprises

income are compiled, when preparing estimates of the extent non-compliance by businesses in- relation to income tax.



National Accounts Branch

-Australian Bureau of Statistics



NATIONAL ACCOUNTS


National accounts are designed to provide a system ie summary of national

economic activity and have been developed to facilitate the practical application of economic theory. At their summary level, the national income, expenditure and product accounts reflect the key economic flows of the Keynesian system: production, income, consumption, investment and saving.


At their more detailed level they are designed to present a statistical picture of the structure of the economy and the detailed processes that make up domestic production and its distribution.


The national accounts show key account aggregates which are generally

accepted as indicators of the state of the economy. Trend in national income and in related aggregates such as gross domestic product give an immediate indication of how-the economy fared in the past year and provide a basis for predicting what is hoped to be achieved in the coming year.


PURPOSES OF NATIONAL ACCOUNTS


Within the constraints of the principles and practice of business and government accounting, national accountants attempt to construct a statistical picture of the key economic processes centred around the measurement of market-based activity (ie activity associated with the production of goods and service resulting in income generation, sales to overseas markets, domestic consumption and capital formation).

National income and other indicators of economic growth measure an important aspect of a country's well-being. However, they do not necessarily provide an indicator of the ultimate (but less quantifiable) objectives relating to the improvement of the welfare or `quality of life' status of the people.


USE OF NATIONAL ACCOUNTS IN DECISION MAKING


National Accounts are powerful tools of policy . In one form or another they

find their way into activities such as the following;







INDIRECT USES OF NATIONAL ACCOUNTS

There are many other areas of application of national accounts in planning and management.


a powerful force in the integration of statistical collections.