Introduction
This publication contains state and territory estimates of gross domestic product (referred to as gross state product (GSP)) and its components, in current price and chain volume terms, for the years 1989-90 to 2020-21. Where ‘state’, is referred to in the text, it also encompasses the two territories. The estimates in this publication are consistent with those published for Australia in the 2020-21 issue of Australian System of National Accounts (cat. no. 5204.0). For further details of the concepts, sources and methods used in compiling the estimates in this publication refer to Australian System of National Accounts: Concepts, Source and Methods (cat. no. 5216.0).
State estimates are essentially a dissection of the Australian estimates contained in Australian System of National Accounts (cat. no. 5204.0). While it is possible in some cases to build up estimates using the same data sources as those used for the Australian estimates, it is quite often necessary to derive dissections using a variety of allocators. These may be directly related to the aggregate being allocated (for example, economic survey data) or only indirectly related (for example, population and household income distributions).
Quarterly state estimates of state final demand (SFD) and its components are released quarterly in the publication Australian National Accounts: National Income, Expenditure and Product (cat. no. 5206.0). The data are available as chain volume measures and in current prices and are presented in trend, seasonally adjusted and original forms.
Concepts, sources and methods
Australia's national accounts statistics are compiled in accordance with international standards contained in the System of National Accounts. These standards are presented in the System of National Accounts, 2008 (SNA08). Australia's application of these SNA standards is described in Australian System of National Accounts: Concepts, Source and Methods (cat. no. 5216.0). It is available on the ABS website http://www.abs.gov.au. This publication outlines major concepts and definitions, describes sources of data and methods used to derive annual and quarterly estimates for major aggregates at current prices and in chain volume terms, and discusses the accuracy and reliability of the national accounts. In addition, it includes documentation on input-output tables, financial accounts, capital stock, productivity measures and balance sheets.
While national estimates are based on the concepts and conventions embodied in SNA08, no such detailed standard is available for sub-national (regional/state) accounts. In the main, the national concepts are applicable to state accounts, but there are a some conceptual differences between national and state accounts. Most problems arise for the Transport, postal and warehousing, Information media and telecommunications, and Finance and insurance services industries, and in the treatment of central government. In such cases, conventions need to be established which reflect data availability and/or the needs of users. Information on the important conceptual, methodological and data issues relating to annual and quarterly estimates by state is provided in Chapter 21 of edition 6 of the Australian System of National Accounts: Concepts, Sources and Methods (cat. no. 5216.0).
Gross state product
The introduction of GSP(P) estimates in 2006-07 resulted in alternative measures of economic activity being available for each state. There were three possible volume measures of GSP which could have been chosen as the headline measure of economic growth:
- volume estimates of GSP(P)
- volume estimates of GSP using the income/expenditure approach (GSP(I/E))
- a simple average of these volume estimates, i.e. GSP(A).
In considering the merits of the various options the ABS decided the average measure is preferred. This measure maximises the use of information about state economic activity and it will be more stable over time (i.e. subject to smaller revisions) than the two alternatives. This approach is also consistent with the approach used nationally for the latest year estimates and for the quarterly national accounts. The volume estimates of GSP(I/E) and GSP(P) therefore contain statistical discrepancies to reconcile them with GSP(A).
The current price estimates of GSP(A) are produced using a simple average of the current price GSP(I/E) and GSP(P) estimates, which are benchmarked to GDP(A). Current price GSP(P) is produced using the production approach where gross value added is equal to output less total intermediate use (GVA = Output - TIU). The current price estimates for GSP(I) and GSP(P) therefore contain statistical discrepancies to reconcile them with GSP(A).
State income estimates
Estimates of GSP(I) in current prices are produced by summing the factors of income (i.e. compensation of employees (COE), gross operating surplus (GOS) and gross mixed income (GMI)) plus taxes less subsidies on production and imports. While wages and salaries can be readily collected by state, this is not always the case for other components of COE nor for GOS plus GMI, which often cannot be measured satisfactorily within business accounting systems at the individual location. Most of the ABS economic collections which provide data for the estimates in this publication obtain various financial information from businesses. Many of these businesses operate within a single state and it is straightforward to attribute the GOS and GMI for those businesses to the appropriate state. For complex businesses that operate in more than one state various conventions have been adopted to allocate GOS and GMI to the appropriate state.
Numerous data sources are used to apportion Australia level GOS to states and territories. Where available, economic survey data, in particular sales data from the Economic Activity Survey, are used to apportion GOS for private non-financial corporations. In the absence of economic survey data, some Australia level industry GOS estimates may be apportioned across states using hours worked estimates from the Labour Force Survey or production-based output estimates. GOS for public non-financial corporations is estimated using Government Finance Statistics. GOS for the general government sector is estimated as equal to consumption of fixed capital estimates derived in calculating capital stock. Similarly, Australia level GMI estimates are apportioned to states and territories in much the same way as GOS estimates. A range of data sources are used, primarily sales data from the Economic Activity Survey. The data sources are selected, wherever possible, so that they represent only unincorporated enterprises.
State estimates of COE by private and public sector are produced consistently with national estimates, compiled for wages and salaries, employers’ superannuation contributions and workers compensation premiums. Australia level industry estimates of COE are apportioned across states and territories using economic survey data, particularly from the Economic Activity Survey and the Survey of Employment and Earnings. The state by industry COE estimates are further constrained by the state by sector COE estimates, to ensure consistency between the two sets of data. There are no individual state by industry COE components of wages and salaries, employers' superannuation contributions and workers compensation premiums. It should be noted that there is a minor difference between COE as published in the total factor income tables and as published in the household income accounts for both the Australian and state estimates. The difference arises in situations where a household supplies labour to a production establishment that is in a different country from the household.
Estimates of taxes less subsidies on production and imports are derived by adding values for taxes received and subsidies paid by state and local governments and the Commonwealth government. Data are available from the ABS Government Finance Statistics (GFS), although Commonwealth government values are not generally available by state. Indicators are used to allocate the goods and services tax, excise tax, customs duties and major subsidies to the state where the economic activity took place.
State expenditure estimates
This publication contains state estimates for a number of expenditure components and international exports and imports of goods and services in both current prices and chain volume terms. The difference between the sum of these components and GSP(E) is known as the balancing item. The balancing item reflects: changes in inventories; interstate trade in goods and services; and the balancing item discrepancy.
Numerous data sources are used to compile the expenditure components both in current price and chain volume terms. For some components state data is available but for others various approaches are taken to allocate the Australia-level estimate to the states. Chapter 21 of the Australian System of National Accounts: Concepts, Source and Methods (cat. no. 5216.0) details the data sources used for the expenditure components of GSP(E).
The chain volume measures of GSP(E) presented in this publication are derived by revaluing current price, income-based estimates of GSP, using deflators which are compiled using the available data on the composition of expenditures on state production and movements in associated prices. For a more complete description of why this approach has been adopted, and details of the data sources and the methods used, see Chapter 21, Australian System of National Accounts: Concepts, Sources and Methods (cat. no. 5216.0).
A lack of data means that it is not possible to derive estimates at the state level equivalent to the aggregate Australian expenditure-based estimates of gross domestic product. While the components of SFD and estimates of international merchandise trade and international trade in services by state are available, there are no complete data on interstate trade in goods and services; or changes in inventories. Hence, an expenditure-based GSP volume measure cannot be derived by a similar methodology to that adopted at the Australian level.
In order to make maximum use of the incomplete expenditure data and limited price data available at the state level, an approach is adopted which essentially involves deriving the best possible current price and chain volume estimates of expenditure that encompass as much as possible of GSP. For each state, current price estimates of identified components of international and interstate trade and changes in inventories are combined with the current price estimates of SFD and, similarly, the chain volume measures of those components are combined with the chain volume measures of SFD. The quotient derived by dividing the aggregate chain volume measure into the aggregate current price estimates is a Paasche (current-weighted) price index. It is this price index which is used to revalue the current price estimates of GSP.
It is crucial to identify separately those components for which the deflators deviate significantly from the average, since the method adopted effectively attributes the weighted average deflator to the unidentified components. There is an assumption underlying the approach adopted in deriving the aggregate state deflators that, because of price competition, the available national price indicators are reasonably indicative at the state level. This is a less distorting assumption if the price indicators are weighted together at a reasonable level of commodity disaggregation, and implies deriving the current price and chain volume estimates at as fine a level as possible.
Despite the fact that as broad a range of information as possible is used in this estimation procedure, the aggregate current price and chain volume estimates of expenditures used in the derivation of the state deflator are not considered to be complete measures. They merely serve to produce the best deflators and therefore the most reliable chain volume measures of GSP(E) that the available data and resources allow.
State production estimates
The same methods currently used to derive Australian level annual volume estimates of industry GVA (for the quarterly and latest year estimates) have been used, where possible, in estimating GSP(P). That is, by using different indicators for each industry to interpolate and extrapolate the supply use benchmark estimates. Most of the indicators are output indicators. One exception is the Agriculture sub-division where a double deflation approach is used. Assumptions have been made, or alternate indicators have been used, on occasions where data availability has limited the application of the national quarterly method.
The GSP(P) method uses an output indicator approach for most industries to compile state by industry GVA estimates. This involves extrapolating reference year estimates of current price GVA using movements in a volume indicator of output. A double deflation methodology is used for the Agriculture sub-division within the Agriculture, forestry and fishing industry.
The method used to derive a current price GVA by state for each industry is to split that particular industry GVA to the states using state based output indicators to split the National output estimates for each industry. National output to TIU ratios are then applied to calculate industry GVA estimates for each state which are benchmarked to the Australian industry GVA estimates.
There are two basic approaches for producing volume indicators, price deflation and quantity revaluation. Some industries only use price deflation while others use a combination of price deflation and quantity revaluation to produce an industry level estimate. Chapter 21 of the Australian System of National Accounts: Concepts, Source and Methods (cat. no. 5216.0) outlines in detail the methods and data sources used for each industry. These two methods provide the output volume indicator which is then used (with corresponding price information) to produce a chain volume measure for each industry.
Once each state's current price and volume GVA estimates have been derived for each industry and Ownership of dwellings they are then benchmarked to the Australian total for each industry. Given the nature of chain volume measures benchmarking results in the volume estimates sum to the Australian total in the reference and current years, but not in other years. Each state's benchmarked industry GVA estimates (current price and other components) are then summed and then re-chained to produce GVA at basic prices for each state. In order to derive GSP(P) for each state, Taxes less subsides on products is added to each state's GVA at basic prices.
Real gross state income
The chain volume measures of GSP measure the volume of goods and services produced in each state. If the terms of trade for a state change significantly (i.e. the prices for a state's exports and imports change at different rates) then chain volume GSP will not accurately reflect the change in real purchasing power of the income generated within a state. For this reason real gross state income (RGSI), has been developed which measures chain volume GSP adjusted for changes in the terms of trade. This measure was introduced in a feature article published in the March quarter 2002 issue of Australian National Accounts: National Income, Expenditure and Product (cat. no. 5206.0).
To obtain an estimate of RGSI, the following adjustment is made to the volume measure of GSP, prior to chaining:
- add exports of goods and services at current prices which are deflated by the implicit price deflator for imports of goods and services;
- deduct the volume estimates of exports;
- the resultant terms of trade adjustment is then added to the volume measure of GSP; and
- the measures of RGSI in the prices of the previous year are then chained to give chain estimates of RGSI.
The estimates of exports and imports of goods and services used in these calculations include both international and interstate trade. The methods used to derive them are described in Chapter 21 of Australian System of National Accounts: Concepts, Source and Methods (cat. no. 5216.0). The international trade data are considered to be of reasonable quality and are presented in this publication. The interstate trade data are derived using a model, and are considered to be of poor quality. Fortunately, the major contribution to changes in the terms of trade of a state comes from its internationally traded goods, principally Australia's exports of primary goods (e.g. minerals) and its importation of manufactured goods (e.g. IT equipment). The contribution of interstate trade to changes in a State's terms of trade is relatively minor because the prices of goods and services exported and imported tend to change at a similar rate.
Accuracy of estimates
The estimates in this publication represent allocations of Australian estimates published in 5204.0. Therefore, they will reflect any inaccuracies in those estimates as well as inaccuracies introduced by any particular conceptual, methodological and data problems inherent in the allocation of Australian estimates to states. The degree of accuracy and reliability will necessarily be lower than that for the Australian estimates.
Uniform methodologies and consistent data sources have been used for all states. Although there is no reason to expect that there would be any bias in the methods used, there is some variation in both the quality of data between states (particularly where sample surveys are used) and the sensitivity of estimates to alternative concepts.
Detailed industry estimates for the two territories are likely to be less accurate than those for the states. The use of indirect allocators to separate the territories has a far more significant effect on the relative errors for the Australian Capital Territory and for the Northern Territory than for New South Wales and South Australia. Where sample surveys are used to collect data, standard errors are often higher for the territories than for the states.
The limitations of ABS sample surveys in producing state estimates are particularly relevant for estimates of private gross fixed capital formation - machinery and equipment. Estimates for each state are derived largely from a quarterly national survey of business. The survey design aims to achieve a high degree of statistical accuracy in the samples in the larger states. However, as the accuracy of the estimates from this survey (in percentage terms) is roughly proportional to the size of the sample, the estimates for the smaller states are generally less accurate than those for the larger states. Hence, the estimates for the Northern Territory, Australian Capital Territory, Tasmania and South Australia are particularly prone to volatile movements.
Estimates of GOS plus GMI for the latest year rely on a range of broad activity indicators such as retail turnover, construction work done, QBIS and persons employed. Therefore, they are subject to revision as more suitable sources, such as taxation statistics and ABS economic survey data, become available.
Estimates of taxes less subsidies on production and imports, government final consumption expenditure, and general government and public corporation gross fixed capital formation can be substantially affected by the indicators chosen to apportion the Commonwealth government component to states. Again, care is required when interpreting these estimates.
In analysing the chain volume measures it is important to recognise the data limitations at the state level and to be aware that the accuracy of the estimates will not be as high as that of the corresponding national estimates.
Other national accounts statistical publications
This publication is part of a regular sequence of national accounts publications. The key national accounts publication is the quarterly Australian National Accounts: National Income, Expenditure and Product (cat. no. 5206.0) released approximately two months after the end of the reference period. The first estimates of national accounts for a particular financial year are released in the September quarter issue of 5206.0, generally in early December. This quarterly publication also contains state estimates for the aggregate State Final Demand and its components.
Also released quarterly is the publication Australian National Accounts: Finance and Wealth (cat. no. 5232.0). It contains financial profiles of each subsector of the economy and the market for each conventional financial instrument.
In late October/early November each year the annual publication Australian System of National Accounts (cat. no. 5204.0) is released. It contains the annual benchmark income, production and expenditure data on which the data in this publication are based. In addition it contains detailed income, capital and financial accounts and balance sheets for all institutional sectors, estimates of productivity and capital stock and a range of industry data. Overall it provides a detailed picture of the structure of the Australian economy.
Input-output tables are available in Australian National Accounts: Input-Output Tables (cat. no. 5209.0.55.001). The latest publication contains input-output tables for 2018-19.
In recent years the ABS has developed estimates of the contribution of tourism to the Australian economy in the form of a tourism satellite account. The latest estimates up to 2019-20 were released in Australian National Accounts: Tourism Satellite Account (cat. no. 5249.0).
Information papers, feature articles and technical notes
Feature articles and technical notes are written on a regular basis to inform users of emerging issues and methodological changes and their impact on the national accounts. Most commonly feature articles and technical notes are released in the quarterly publication Australian National Accounts: National Income, Expenditure and Product (cat. no. 5206.0) but other publication vehicles are also used. A full list of feature articles published since December 1988 is included on the ABS website.
The ABS website includes a National Accounts theme page which contains links to the latest National Accounts releases, feature articles and information papers. It also contains links to other related sources of information including non-ABS sources.