Australian Bureau of Statistics
5204.0.55.009 - Information Paper: Australian National Accounts, Distribution of Household Income, Consumption and Wealth, 2009-10
Latest ISSUE Released at 11:30 AM (CANBERRA TIME) 16/08/2013 First Issue
|Page tools: Print Page Print All RSS Search this Product|
The very remote communities and people living in non-private dwellings, populations that were out of scope of the micro surveys (1% and 1.5%, respectively of the resident population), were excluded from the ASNA estimates and distributed separately using data from the 2011 ABS Census of Population and Housing. These distributions were then added to the ASNA distributions based on the micro surveys to obtain the final distribution of the ASNA household income, consumption and wealth estimates.
MACRO DATA SOURCES - ASNA HOUSEHOLD ACCOUNTS
Current price estimates for income, consumption and wealth were sourced from the household sectoral tables: income account; final consumption expenditure; adjusted disposable income account; and balance sheet from the 2011-12 issue of the Australian System of National Accounts (cat. no. 5204.0). Appendix 2 provides the ASNA components of income, consumption and wealth that were distributed. ASNA household estimates measure the resident population living in private dwelling households, institutional households (e.g. long term patients in hospitals and people in retirement homes), unincorporated enterprises and non-profit institutions serving households (NPISH). NPISH estimates are included in the ASNA household sector due to the lack of sufficient data to enable the construction of a full range of sector accounts for NPISH activities. For an explanation of the concepts, sources and methods for the ASNA items see Australian System of National Accounts, Concepts, Sources and Methods, Australia 2012, Edition 3 (cat. no. 5216.0).
MICRO DATA SOURCES - ABS HOUSEHOLD SURVEYS
Micro information was sourced from the cross sectional 2009-10 ABS Survey of Income and Housing (SIH) and 2009-10 Household Expenditure Survey (HES). The SIH and HES collect information from individual households which was used to formulate distributional information of income, expenditure and wealth across the population and between various population subgroups. Both surveys collect detailed information about income, assets, liabilities and household characteristics of persons aged 15 years and over resident in private dwellings throughout Australia, however the main focus of the HES is consumption expenditure.
The study also utilised information from four household publications based on the SIH and HES:
The concepts and definitions relating to income, consumption, net worth, government benefit and taxes and households are included in the explanatory notes and glossary in the above publications. The survey methodology and reliability of estimates for SIH are described in explanatory notes in 6523.0 and 6554.0 and for HES in 6530.0
The household surveys exclude part of the population that was covered in the ASNA, the survey out of scope population relate to those individuals living in non-private dwellings (such as hotels, boarding schools, boarding houses and institutions), and those living in very remote communities. In the 2009-10 surveys, the out of scope population was approximately 3% of the resident population.
The SIH and HES do not provide information for a number of items that are part of the national accounts definition of income, consumption and wealth, for example, employer imputed social contributions, financial intermediation services indirectly measured (FISIM), property income attributed to insurance policy holders and fixed assets such as cultivated biological resources. These components are missing from household surveys due to conceptual and practical reasons. For example, compilers of micro statistics consider that some national accounts components that are useful to describe the economy as a whole, such as FISIM, are not relevant when the focus is the economic behaviour of households. On the other hand, property income attributed to insurance policy holders is missing from micro estimates because the information is difficult to collect or impute.
The income estimates from the micro sources included in the information paper include imputed rent for owner occupied housing and subsidised private rentals. This aligns with the definition of income included in ABS cat. no. 6537.0, but is broader than the definition used in the other household survey publications (e.g. cat. no. 6523.0). The micro income estimates include the following components that are not in scope of the ASNA, income from superannuation, annuities, private pensions and income from family members not living in the household, and excludes, income in scope of the ASNA such as imputed property income on investment income earned by insurance policy holders and financial intermediation services indirectly measured (FISIM). As a consequence, the income variable used to classify households into groups are not fully consistent with the national account income aggregates.
Divergences are also observed between survey and national accounts estimates because the micro surveys follow different classification rules to the national accounts for equivalent estimates. For example, wages and salaries reported in the surveys include workers' compensation received through the payroll, where as workers' compensation payments are a separate item in ASNA income accounts; unlike the survey estimates, dividends in the ASNA do not include imputation credits. However, for consumption estimates used in this information paper, there was generally good alignment between the micro and macro estimates. The initial HES data is coded to a detailed household expenditure classification (HEC), and then concorded to the United Nations' Classification of Individual Consumption by Purpose (COICOP) which is the classification used in ASNA. Other divergences that occur between micro data and national accounts reflect differences in valuations, for example imputed rent for owner occupied housing and, statistical adjustments applied, for example for non-response and sampling error.
To distribute the national account estimate by the five household groups (main source of income, equivalised income quintiles, household composition, age of household reference person, and equivalised quintiles) five steps were followed.
Step 1: Scope adjustment to national account aggregates and components
The ASNA household income, consumption and wealth estimates from 5204.0 for 2009-10 were adjusted to remove transactions related to NPISH, people living in non-private dwellings (such as nursing homes, hotels, boarding houses and institutions) and very remote communities. The household micro data used for distribution into household groups does not include these populations, and it was assumed that this out of scope population have income, consumption and wealth patterns that differ from those of the population covered by micro sources.
Non-profit institutions serving households (NPISH)
The NPISH income component, current transfers to non-profit institutions from general government and public corporations were directly removed from the household income account. The NPISH estimates for interest, dividends, income tax and other current transfers were modelled for 2009-10 from estimates in the Australian System of National Accounts: Non-Profit Institutions (NPI) Satellite Account, 2006-07 (cat. no 5256.0). The modelled NPISH estimates were removed from the income estimates in the household income account.
The ASNA household consumption estimates for education, health, transport and accommodation services were adjusted for NPISH. The estimates for current grants from general government were sourced from ABS Government Finance Statistics and the estimates for grants and donation from corporations and households were modelled from 2006-07 NPI satellite account. The final consumption estimates taken out for NPISH were then included in the household adjusted disposable income account as social transfers in kind provided by NPISH for education, health and other services.
The 2009-10 NPISH assets and liability estimates were modelled on the 2006-07 NPI satellite account, and experimental information on NPISH non-financial assets published in the ABS, Working Paper in Econometrics and Applied Statistics, Working Paper No. 2002/1, Experimental Estimates of the Distribution of Household Wealth, Australia, 1994-2000 (cat no. 1351.0). The modelled estimates were removed from the household balance sheet.
People living in non-private dwellings and very remote communities
For people living in non-private dwellings (NPD), and very remote communities (VRC), income, consumption and wealth components were calculated using demographic and economic information from the 2011 ABS Census of Population and Housing (2011 Census). It was assumed that the economic and demographic structure of the Australian population would have experienced minimal change since the financial year 2009-10 up to August 2011 (month the 2011 Census was undertaken).
The out of scope population for VRC were split into indigenous and non-indigenous communities. The NPDs were separated into three categories, institutions related to health and aged care, welfare institutions, and other. Corrective institutions such as prisons were excluded from the NPD estimates as the census data recorded minimal to nil income for inmates; the consumption expenditure for inmates were classified in ASNA as government final consumption expenditure; and it was assumed that their balance sheet (i.e. wealth) would be insignificant.
Shares of the NPDs and VRC (to the whole of Australia) were calculated from 2011 Census data, for population, employment status, age profile, and home ownership. The estimates were then used to calculate out of scope income, consumption and wealth components that were removed from the national accounts aggregates.
The following information from 2011 Census were used for VRC:
The following information from 2011 Census were used for NPDs:
Step 2: Comparison of macro and micro estimates
The national accounts household income, consumption and wealth estimates (macro) after step 1, was compared to the corresponding household survey estimate (micro). Raw coverage ratios, that is the micro estimate divided by the macro estimate as a percentage were then calculated to assess the correlation between the two estimates. For some items, 'adjusted' coverage ratios were calculated, for these components further adjustments were made to macro or micro (or both) estimates to enable the most relevant common scope for comparison. The 'raw' and 'adjusted' coverage ratios are published in electronic table 9. A coverage ratio between 75% to 125% indicates good alignment between the micro and macro estimate, and therefore direct distribution using the micro distributional household indicator would be the chosen methodology. For example, the adjusted coverage ratio for social assistance benefits was 84% and was directly distributed to the household groups using the survey micro distributional information. However, bias could occur if under coverage or other reporting errors differ between household groups in the distributional indicator.
The ABS publications based on 2009-10 SIH and HES publish a comparison between the ASNA and corresponding survey estimate for income (Appendix 5, cat. no. 6523.0), consumption (Appendix 3, cat. no. 6530.0) and wealth (Appendix 3, cat. no. 6554.0). The ASNA estimates in the appendices are from the 2009-10 release of cat. no. 5204.0 (released on 28 October 2011) and are not adjusted for populations not in scope of the surveys and some NPISH estimates are still included. The coverage ratios published in these publications are marginally different to those presented in electronic table 9 in this information paper. Despite these differences, the appendices should be referenced for a detail description of the main scope, definitional and methodological difference between the micro and macro estimates .
The corresponding micro estimates to the ASNA income, consumption and wealth estimates are sub sectored into five household groups: main source of income; equivalised income quintiles; household composition; age of household reference person; and equivalised net worth quintiles. The household groups are created using information available at the individual and at the household level in micro sources. The criteria pursued in the choice of the households groups selected for this information paper were to:
The groups selected are also considered useful for economic analysis and policy purposes.
Main Source of Income (MSI)
Households are classified according to the main source of income for the household as a whole. The five income sources identified were wages and salaries, income from unincorporated business, property income and superannuation, government pensions and allowances, and other. The MSI categories selected in this information paper are different to the MSI categories published in the ABS micro publications 6523.0, 6530.0, 6554.0 and 6537.0. The differences are: micro publications do not separate the MSI category 'property income and superannuation', they are included in the micro category 'other'; the micro category for MSI 'own unincorporated business income' does not include imputed income from owner occupied dwelling, those households are included in micro category 'other'. Table 4.1 below illustrates the major micro components, number of households and value in each category of the MSI used in this information paper, the information in column 2 to 4 are from ABS micro sources.
Table 4.1, Components of Main Source of Income
Equivalised Income Quintiles
Households were classified according to the level of their equivalised disposable income (EDI) as it enabled comparison of the relative economic well-being of households of different sizes and composition. EDI was calculated by adjusting disposable income by the 'modified OECD' equivalence scale. The scale assigns a value of 1 to the household head, 0.5 to each additional adult member and 0.3 to each child under 15. The scale reflects the requirement for a larger household to have a higher level of income to achieve the same standard of living as a smaller household. When household income is adjusted according to the equivalence scale, the equivalised income can be viewed as an indicator of economic resources available to a standardised household. For a single person household, it is equal to income received. For a household comprising more than one person, equivalised income is an indicator of household income that would be required by a single person household in order to enjoy the same level of economic well-being as the household in question. To obtain EDI quintiles, households were ranked according to the value of the equivalised disposable income and grouped into quintiles, each quintile comprised the same number persons, that is the quintiles were person weighted. For more information on the use of equivalence scales, see Appendix 3 in Household Income and Income Distribution, Australia, 2009-10 (cat. no 6523.0).
Households were classified according to three criteria: the number of adults in the household; the age of the adults; and the presence of children living at home. Seven household sub-groups are distinguished:
Households are classified according to six household sub-groups by age in the following years:
The reference person for each household is chosen by applying to all household members aged 15 years and over, the selection criteria below, in the order listed, until a single appropriate reference person is identified:
Equivalised Net Worth Quintiles (ENW)
Household equivalised net worth is calculated by adjusting net worth by application of the 'modified OECD' equivalence scale. The same methodology is applied to calculate equivalised net worth as described for equivalised disposable income quintiles. To obtain ENW quintiles, households were ranked according to the value of their equivalised net worth and grouped into quintiles, each quintile comprises the same number persons, that is the quintiles are person weighted.
It is generally accepted that the appropriate presentation of income quintiles should be in terms of equivalised income. However, there is far less agreement about delineating wealth quintiles in equivalised terms. For example, the new guidelines for household wealth released by the OECD do not provide a definitive recommendation on equivalising wealth. The following excerpt is taken from the OECD Guidelines for Micro Statistics on Household Wealth, 2013, Chapter 7.3.6 on equivalence scale.
In the case of household income, there are internationally recognised equivalence scales that are used to standardise the estimates with respect to household size and composition while taking into account the economies of scale that arise from living together, in particular through the sharing of dwellings. In the case of household wealth, however, no internationally agreed equivalence scales exist, and there is no consensus on whether the scales used for income are appropriate for wealth. In studies jointly analysing income and wealth, the equivalence scale applied to income is also applied to wealth (OECD, 2013).
The use of equivalence scales in the case of wealth depends on the purpose of the analysis. Equivalence scales should not be used when analysing the characteristics of individual components of wealth. If, on the other hand, wealth is treated as a source of income streams that can be used to finance consumption and contribute to economic well-being in the household, wealth might be equivalised just as income. Equivalised estimates are often expressed in terms of single-person household equivalents (i.e.. the level of wealth that would be required by a lone person household to have the same level of economic well-being as the household in question). Failure to equivalise could provide a misleading picture of the distribution of wealth, for example by overstating the share of single-person households at the bottom of the distribution.
In this information paper, equivalised net worth was used as the preferred household distributional indicator as the paper jointly analyses and presents household quintiles for income and wealth. A major argument for equivalising household wealth is that it does affect current economic well-being in some circumstances as it provides a buffer if income streams are disrupted and is likely to affect consumption decisions. However, it is recognised that wealth is generally accumulated over the working life of households so the end beneficiaries are not necessarily those present in the current household and therefore the use of the income equivalisation scale for wealth is not as appropriate when analysing long term beneficiaries of wealth.
Step 4: Allocation of the scope adjusted national accounts estimates to the five household groups
The scope adjusted national accounts estimates for income, consumption and wealth were distributed using the micro distributional information of the five household groups by:
Tables 4.2, 4.3, 4.4 and 4.5 below summarise the methodology used to distribute ASNA income, consumption and wealth estimates.
Table 4.2, Income Accounts - Income
Table 4.3, Income Accounts - Final Consumption Expenditure
Table 4.4, Adjusted Disposable Income Account
The micro distributional components for social in transfers in kind were modelled from survey and other data sources such as ABS Government Finance Statistics, for further information on the micro distributional components please refer to cat. no. 6537.0.
Back to top
Table 4.5, Balance sheet
Step 5: Distribution of the out of scope population.
For people living in non-private dwellings (NPD), and very remote communities (VRC): the disposable income (derived from Step 1) was 0.7% and 0.8% of total household disposable income; 1.0% and 1.1% of total consumption expenditure; and 0.8% and 0.6% of total net worth. As discussed in Step 1, it was assumed that the distribution pattern for NPD and VRC would differ from the population covered by micro sources and therefore separate distributions for the five household groups were derived to add back into the macro distributions derived from the micro surveys. Despite the contribution of these population groups being approximately only 1% to the national account aggregates such as disposable income, it is important to include these distributions as the aim of this study is to distribute fully the ASNA household estimates.
Information from the 2011 Census (specified in Step 1) are used to derive economic and demographic profiles for the NPD and VRC categories and the whole of Australia.
The following profiles and the share of the population and households in each profile are derived:
The information from the profiles above indicated that the non-indigenous VRC were very similar to Australia as a whole, therefore their income, consumption and wealth estimates were distributed on the basis of the micro data (as described in Step 4).
The census profiles for indigenous VRC, health and aged care institutions, welfare institutions and other NPDs are used to distribute the national accounts 'out of scope' estimates (calculated in Step 1) into the five household groups. Distribution for some components required additional assumptions and imputation as detail data required was not available. For NPDs, it was assumed that the individual population count equated to equivalent number of households.
The methodology used in this study to allocate the national accounts totals raised the following methodological issues:
(1) The household distributions were estimated for income, consumption and wealth using both the SIH and HES. The income and consumption components were then combined to estimate the national accounts aggregates, e.g. adjusted disposable income, actual final consumption and saving by household groups. The approach assumed that for each household group the two surveys used describe on average, the same type of households. If in reality, the households in the SIH were different from the households in the HES, there would be an impact on the accuracy of the derived national accounts aggregates such as gross saving. Further, some irregularly purchased items (for example, motor vehicles) in the HES can be impacted by small variations in the household sample that reported such expenditure, particularly in analysis of sub-populations. This in turn would have impacts on any derived national accounts aggregates.
(2) The robustness of the macro distributional information derived from the micro sources depends on the sample sizes of the surveys, when disaggregated survey estimates were used, in some instances, the micro distribution utilised for household groups had significant standard errors.
(3) Methods B and C described in this chapter, under section 'The distributional methodology of the household national accounts' step 4, were based on assumptions about similarities in distributional patterns between macro components. In some cases, the assumption may not be robust but were applied as a 'best' method available due to lack of source data. This issue was identified in particular at the OECD/EuroStat expert group, where some country experts did not support the distribution of imputed national account estimates such as FISIM, due to the appropriateness of indicators available for such distributions, and also the validity of the distribution an imputed estimate was questioned.
(4) The distribution of the out of scope population from micro surveys as described in chapter 3, under section 'The distributional methodology of the household national accounts' steps 1 and 5, were based on distributions derived from the 2011 census data. The derivations of these distribution were based on assumptions placed on the census data and in some cases when no census data was available, best estimates were imputed. The issue of the robustness of the applied models for some out of scope estimates were questioned. In particular, the assumption that the NPD individual equates to equivalent number of households. The overall addition of the number of households to the micro survey data due to the addition of including the out of scope population was approximately 4%, the biggest driver for the increase was the NPD household estimate. The NPD household count assumption was made as no information or methodology was available to distribute this population in any alternate way. It was also recognised that there would be a minor impact on the original equivalised distribution patterns provided by the micro sources due to the addition of the extra households into equivalised income and net worth micro distributions (see (5) below). However, the impact of the models and the underpinning assumptions for out of scope population was considered minimal on the overall distribution of the household national account estimates as the out of scope population impact on the national accounts income, consumption and wealth aggregates was around 1%.
(5) For both macro and micros sources, the preferred unit of analysis is the household. Since equivalised household income and net worth indicators can be viewed as measures of the economic resources available to each person in a household, their estimation from micro data sources in this publication is based on numbers of people rather than households. This is known as person weighting and ensures that people in large households are given as much weight in the distribution as people in small households. This means that the lowest income quintile represents the 20% of people in households with the lowest household income. For national accounts purposes it would be preferred that quintiles were based on equal proportions of households rather than persons, that is, the lowest income quintile represents the 20% of households with the lowest household income.
Table 4.6 below, shows the percentage of persons and households, in the equivalised income and net worth quintiles used in this study. The percentages of households in each quintile are shown, directly from micro sources and after the out of scope micro population is added.
Table 4.6, ASNA, Shares, equivalised income and net worth quintiles, 2009-10, per cent
While the table indicates that the concordance between income quintiles for persons and households is generally quite close, the net worth quintiles for persons and households display slightly larger differences. In this study we recognise that the equivalised net worth quintiles while analytically useful do have some methodological issues that need to be addressed; please refer to the discussion on the derivation of the net worth quintiles (step 3 of this chapter).
These documents will be presented in a new window.
This page last updated 15 August 2013