### Introduction

This publication contains estimates of gross domestic product (GDP) and its components, components of state final demand, the national income account, the national capital account and supporting series. For the most part, these estimates are provided in trend and seasonally adjusted terms. Where trend and seasonally adjusted estimates are not available, original data are provided. Annual estimates, on an original basis, are also provided for the key aggregates. The list of time series spreadsheets (below) shows the full range of data provided.

### Concepts, sources and methods

Australia's national accounts statistics are compiled in accordance with international standards contained in the System of National Accounts, 2008 (2008SNA). Australia's application of these SNA standards is described in Australian System of National Accounts: Concepts, Sources and Methods (cat. no. 5216.0). This publication outlines major concepts and definitions, describes sources of data and methods used to derive the full suite of national accounts estimates for Australia.

The system could also be defined more widely to include balance of payments and government finance statistics. For detail see Balance of Payments and International Investment Position, Australia: Concepts, Sources and Methods (cat. no. 5331.0) and Australian System of Government Finance Statistics: Concepts, Sources and Methods (cat. no. 5514.0).

While national estimates are based on the concepts and conventions embodied in 2008SNA, no such standard is available for sub-national (regional/state) accounts. In the main, the national concepts are applicable to state accounts, but there remain a number of conceptual and measurement issues that either do not apply or are insignificant at the national level. Information on some of the more important conceptual, methodological and data issues relating to annual and quarterly estimates by state is provided in Chapter 21 of the Concepts, Sources and Methods.

### Gross Domestic Product (GDP)

GDP is derived by three approaches: the income approach (I), the expenditure approach (E) and the production approach (P). A description of each approach is provided in the following paragraphs. While each measure should, conceptually, produce the same estimate of GDP, if the three measures are compiled independently using different data sources, then different estimates of GDP result.

The ABS aligns the estimates of GDP annually by balancing them in supply and use tables. These tables have been compiled from 1994-95, up to the year preceding the latest complete financial year. Balancing in supply and use tables ensures that the same estimate of GDP is obtained from the three approaches. Annual estimates using the I, E and P approaches are identical for the years for which these tables are compiled. For years balanced using supply and use tables, quarterly GDP is benchmarked to annual GDP. However, the three estimates of GDP can be different for any given quarter. The annual GDP estimate produced by balancing using supply and use tables forms the benchmark for the production of quarterly GDP going forward. Quarterly GDP is compiled in chain volume terms using all three approaches. The headline measure of GDP is a simple average of the three separate measures. It is labelled GDP(A), with "A" denoting "average".

Prior to 1994-95 quarterly and annual estimates using each approach are based on independent sources, and there are usually differences between the I, E and P estimates. Nevertheless, for these periods, a single estimate of GDP has been compiled. In chain volume terms, GDP is derived by averaging the chain volume estimates obtained from each of the three independent approaches. The current price estimate of GDP is obtained by reflating the average chain volume estimate by the implicit price deflator derived from the expenditure-based estimates.

As a result of the above methods:

- There is no statistical discrepancy for annual estimates from 1994-95 up to the year prior to the latest complete financial year, in either current price or volume terms, except for estimates released in the June quarter where discrepancies will exist for the latest two complete financial years.
- For years prior to 1994-95, and for all quarters, statistical discrepancies exist between estimates based on the I, E and P approaches and the single estimate of GDP, in both current prices and volume terms. These discrepancies are shown in the relevant tables.

#### Income approach (I)

GDP using the income approach is derived as the sum of compensation of employees, gross operating surplus, gross mixed income and taxes less subsidies on production and imports. Volume estimates are derived at the total GDP level by deflating current price estimates by the implicit price deflator from the expenditure approach.

#### Expenditure approach (E)

GDP using the expenditure approach is derived as the sum of all final expenditures, changes in inventories and exports of goods and services less imports of goods and services. Volume estimates are derived for each of the components as well as for their sum.

#### Production approach (P)

GDP using the production approach is derived as the sum of gross value added for each industry, at basic prices, plus taxes less subsidies on products. Basic values represent the amounts received by producers, including the value of any subsidies on products, but before any taxes on products. The difference between the sum over all industries of gross value added at basic prices, and GDP at market (or purchasers') prices, is the value of taxes less subsidies on products.

In this publication, only volume estimates compiled using the production approach have been shown. These estimates are derived by extrapolating annual volume measures using various indicators. The information necessary to compile comprehensive current price estimates using the production approach is not available quarterly.

### Accuracy of quarterly estimates

Estimating the national accounts components for a period of less than one year presents special problems. It is often difficult to adhere strictly to definitions used in annual estimates when deriving quarterly ones. This is particularly the case for the quarterly measure of income, because it is not always possible to match the volume of production for a quarter with the cost incurred in that production. Difficulties are also encountered in obtaining detailed data for short periods and in preparing consistent estimates from different sources with different accounting procedures and periods. Furthermore, the quarter-to-quarter growth in seasonally adjusted terms is very sensitive to the timing of recording a transaction. If the recording of a transaction is delayed by one quarter, seasonally adjusted movements will be distorted for three consecutive quarters. All these problems affect the accuracy of the current price and chain volume estimates and should be taken into account in interpreting the estimates.

The majority of the estimates in the quarterly national accounts are based on the results of sample surveys. Many of the results of these surveys are released in the period leading up to the release of the quarterly accounts, thus providing users with a guide to likely movements in key national accounting aggregates. In a national accounts context, these various pieces of information are referred to as partial indicators. Usually there are differences in concept and scope between the national accounts series and the relevant partial indicator which means that the movements in the partial indicator will not always be identical to the national accounts series movement. However, in general the movements should be similar. To ensure a reasonable level of consistency between the partial indicators and the national accounts series and hence present a common understanding of recent economic developments, the national accounts area liaises with the relevant survey areas and provides feedback regarding data quality and data coherence. This may result in adjustments being made by survey areas to their collected data prior to their release. The objective use of the national accounts framework to provide data coherence across all ABS economic statistics ensures that a common understanding of recent economic developments is presented.

The state estimates generally represent dissections of quarterly estimates published for Australia in this publication. Therefore, they will reflect any inaccuracies in those estimates as well as inaccuracies introduced by the particular conceptual, methodological and data problems inherent in the allocation of Australian estimates to states and territories. As such, the degree of accuracy and reliability will necessarily be lower than that for the counterpart Australian estimates.

Estimates for compensation of employees, household final consumption expenditure and private gross fixed capital formation are based on the results of sample surveys. By their nature, survey results become less accurate as they are disaggregated (for example into states and/or industries). Generally, the ABS surveys used to derive these aggregates are designed to provide accurate estimates at the Australian total level and individual state estimates that are less accurate but still of acceptable quality. However, it should be noted that relative standard errors are generally higher for the smaller states and territories than they are for the larger states. This may result in greater volatility in the quarterly estimates for the smaller states and two territories.

Estimates of government final consumption expenditure, and general government and public corporations gross fixed capital formation can be substantially affected by the indicators chosen to apportion the Commonwealth government component to states and territories. Care is required when interpreting these estimates.

### Seasonally adjusted and trend estimates

The general methods for deriving seasonally adjusted and trend estimates are described in Australian System of National Accounts: Concepts, Sources and Methods (cat. no. 5216.0).

For more information about time series analysis, see Time Series Analysis Frequently Asked Questions.

#### Seasonal adjustment

Data that are affected by seasonal factors are adjusted to remove the effects of these factors. Three important points should be noted here:

- The methods used in seasonal adjustment do not force the sum of the adjusted current price estimates for each quarter of a year to equal the original annual total.
- Where chain volume estimates have no apparent seasonality in their implicit price deflators, the estimates are adjusted using the corresponding factors for current price estimates.
- A special method, known as the pseudo-additive method, has been used to adjust the output of cereal crops. This is necessary to account for the fact that there is no cereal output in some quarters.

Seasonally adjusted chain volume figures are calculated from seasonally adjusted figures expressed in the prices of the previous year. As with original data, the seasonally adjusted chain volume measures are benchmarked to annual original estimates. As a consequence, the seasonally adjusted chain volume measures sum to the corresponding annual original figures - unlike their current price counterparts.

#### Trend estimates

A trend estimate is obtained by removing the irregular component from the seasonally adjusted series. For estimates in this publication, it is calculated using a centred 7-term Henderson moving average of the seasonally adjusted series. The procedure is designed to minimise distortions in the trend level, turning point shape and timing of turning points. Estimates for the three most recent quarters cannot be calculated using this centred average method; instead an asymmetric average is used. This can lead to revisions in the trend estimates for the last three quarters when data become available for later quarters, even if none of the original data for earlier quarters has changed.

The higher the 'irregular' component in a series, then the greater the likelihood that trend estimates for the latest quarters will be revised as more observations become available. However, it is important to note that this does not make the trend series inferior to the seasonally adjusted or original series. In fact, in such cases the effect of the irregular component on overall movements is likely to be even more in the seasonally adjusted and the original estimates than in the trend series.

Trend estimates for aggregates such as GDP are derived directly, rather than as the sum of components. As a result, the sum of the trend estimates of individual components of a particular aggregate will not sum to the overall trend estimate of the aggregate for the latest three quarters. This approach provides higher quality trend estimates for key aggregates, particularly GDP.

For more information about ABS procedures for deriving trend estimates and an analysis of the advantage of using them over alternative techniques for monitoring trends, see Information Paper: A Guide to Interpreting Time Series - Monitoring Trends (cat. no. 1349.0) or contact Time Series Analysis by email at time.series.analysis@abs.gov.au.

#### State and territory versus Australian series

For trend and seasonally adjusted series, the sum of the states and territories generally does not equal the corresponding estimate for 'total Australia', nor are the quarter-to-quarter movements identical. On a few occasions, these differences have been significant, particularly for the seasonally adjusted series. This reflects both the shorter span of data available for seasonal analysis at the state level and the fact that seasonal analysis is generally carried out at a more aggregated level than for the 'total Australia' series. The state and territory trend and seasonally adjusted series are less accurate than the Australian data. However, as the state and territory time series lengthen, the quarterly movements in the sum of the state estimates should more closely match those in the Australian series.

### Revisions

Most figures are subject to revision as more complete and accurate information becomes available. The revisions are of two types: those made to recent quarters and those made as a consequence of a redistribution across all quarters within a year following revisions to annual totals.

### Related publications

In addition to the publications already mentioned, others of interest include:

- annual Australian System of National Accounts (cat. no. 5204.0);
- annual Australian National Accounts: State Accounts (cat. no. 5220.0); and
- quarterly Australian National Accounts: Finance and Wealth (cat. no. 5232.0).

Current publications and other products released by the ABS are freely available from the ABS website http://www.abs.gov.au.