Revisions

24.107    Revisions in relation to the national accounts may be defined as the differences between a published sequence of estimates for a given reference period for the same national accounting variable over a particular period of time. Revisions arise from the progressive incorporation of more up to date data, re-weighting of chain volume series and reassessment of seasonal factors, and from time to time the introduction of new accounting concepts and improved data sources and methods.

24.108    Revisions analysis is concerned with the reliability of initial (or subsequent) estimates rather than the accuracy of estimates. Accuracy is always the main focus of statistical agencies. However, a standard accuracy measure, such as the standard error of GDP, has been found impossible to produce in practice because of the complicated compiling process and aggregation structure of the national accounts. For example, a Eurostat (1999) task force concluded that:

. . . even under the most optimistic assumption that confidence intervals can be provided for all basic sources, objective error margins for national accounts aggregates appear to be out of reach. The problems posed by identifying, measuring and aggregating all errors that remain after adjustments are made at the various compilation stages seem insurmountable.

24.109    Reliability refers to the ability of the successive vintages of national accounts estimates to present a consistent, reliable picture of the economy, as estimates are revised to incorporate increasingly comprehensive and improved source data. It is therefore a relative measure rather than an absolute one. Revisions analysis is often employed as a diagnostic tool to indicate possible sources of unreliability, but the results should never be used as an adjustment factor in compiling or interpreting the national accounts.

24.110    Revisions should not be seen as synonymous with error. They are an inevitable consequence of the national accounts process. Revisions reflect both the complexity of economic measurement, and the need to provide economic policy advisers and other users with timely estimates, in order to maximise their use in analysis of current economic conditions. A major reason for revisions to initial estimates is the trade-off between timeliness, accuracy and reliability. For example, delaying the release of data by a year or more would not mean that revisions are no longer required. There are several reasons:

24.111    The national accounts process is aimed at bringing together sets of accounts that are coherent for the whole nation, unlike in business accounting, where the aim is to record accurately the transactions and financial position of a particular business entity in isolation. It is, in effect, a quadruple accounting system, and achieving coherence between the transactions of economic agents is an iterative process proceeding over a number of years.

24.112    Some of the most complete data are only available every few years (such as the Census of Population and Housing).

24.113    Seasonally adjusted and trend estimates are updated as information from new quarters is used to improve estimates of the trend, seasonal and irregular components of each time series.

24.114    Improvements to data sources and estimation methods are backcast to earlier periods where possible, as well as occasional changes to national accounting standards.

24.115    ABS and international data quality assessment frameworks include revisions history as one of the indicators of quality, linking it with the accuracy element. Users need to be aware of the potential for revisions. They need to satisfy themselves that initial estimates provide an acceptable indication of later estimates based on more complete data, which has been subjected to coherence checks within the national accounts system. As part of an overall quality assessment program, national accounts compilers are encouraged to regularly conduct an analysis of revisions in order to inform users, and to help guide the continuous quality improvement process within statistics agencies.

24.116    The ABS has published studies of revisions in the past, as have a number of other national statistics agencies. Since 2005, the OECD have occasionally published data on revisions to quarterly GDP and other economic indicators in selected countries. The database has been designed for a limited number of key economic statistics together with an analytical spreadsheet. It is also designed to assist compilers and analysts in the derivation of a standard set of revisions measures, using either OECD data or their own data. The database facilitates international comparisons of revisions.

24.117    The remainder of this section describes:

1. the national accounts compilation process;
2. the revisions policy; and
3. the availability of information on revisions.

24.118    The ABS has implemented a set of rules to manage the timing of revisions to the national accounts. These rules are based around the national accounts compilation process, which itself reflects the availability of new or revised source data, as well as operational factors.

24.119    The basic accounting and price index data from the quarterly surveys can normally be expected to become quite firm the quarter following the initial estimate. The national accounts process requires that the estimates derived from these sources be subject to annual benchmarking. The results of the benchmarking process are first published in the September quarter issue of the quarterly national accounts, coinciding with the availability of balanced annual current price and volume data from the Supply and Use tables.

24.120    The Supply and Use system progressively incorporates business accounts and other data from ABS annual surveys, as well as the taxation system. A balancing process is undertaken to achieve consistency between the supply and use of products in the economy, in both current price and volume terms. Data inconsistencies are reviewed and have to be resolved by altering some of the basic data. Quarterly national accounts estimates are benchmarked to successive vintages of annual data in order to maintain consistency within the national accounts system. This process also introduces revisions to quarterly growth rates in years either side of the new or revised annual data.

24.121    The benchmarking procedure used by the ABS is a mathematical routine designed to equate the quarterly and annual benchmarked estimates, while minimising the impacts on existing quarterly growth rates. Using this procedure, revisions to an annual estimate have the potential to impact on growth rates for all the quarters in that year, in the two previous years and two forward years. Impacts of benchmarking on individual quarter growth rates could be reduced by adopting a simple pro rata procedure, but this would force all of the revision to growth rates into the September quarter. The optimising procedure used is widely considered to produce superior estimates of quarterly growth rates, but at the expense of more widespread revisions. Revisions to annual estimates result from the progressive incorporation of more complete source data balanced in the Supply and Use system, and revisions to quarterly estimates result from the process of benchmarking. In this way, a revisions process is inherent within the national accounts process because more comprehensive and reliable data only become available with a considerable lag.

24.122    Typically, the national accounts process for a given quarter for original current price and volume data proceeds as follows:

• The initial quarterly estimate is based on preliminary quarterly survey data.
• The later vintage quarterly estimates up to the June quarter issue of the national accounts are based on more complete or 'final' quarterly survey data.
• At the end of the current financial year (June quarter) the annual estimates for that year are derived initially as the sum of the four quarters.
• In the September quarter, the reference year values for the whole time series of chain volume estimates are advanced one year to the current price values applying in the previous financial year. Re-referencing impacts data levels, but not growth rates. The base year weights for the current and previous year chain volume estimates are also moved forward. This can result in a revision to growth rates in those years.
• In the September quarter, the quarterly estimates are subject to a benchmarking process to align them with annual current price and volume data that has been balanced in the supply and use system.

24.123    As noted in the Methodology subsection above, the X-11 seasonal adjustment approach used by the ABS decomposes original current price and chain volume estimates into a combination of modelled trend, seasonal and irregular components, which are used to produce seasonally adjusted and trend estimates. The addition of new quarters and data revisions to previous quarters results in a re-estimation of the seasonal factors and underlying trends in the data due to the end-point problem. Revisions to growth rates will result from this process. The usual concurrent seasonal adjustment method involves re-analysing and re-estimating seasonal factors for the whole time series with the addition of each new quarter, and, therefore, revisions to seasonally adjusted and trend estimates potentially flow through with the addition of each new quarter. These revisions tend to be largest over the most recent few years. Components for which seasonal factors are re-analysed and re-estimated annually will show revisions at the time this occurs. In other quarters, the latter approach will not revise seasonal factors, but any revisions to original data may still flow through to seasonally adjusted and trend estimates.

24.124    As a consequence of this process, the following revisions policy applies for GDP and components in normal circumstances.

24.125    The September quarter releases normally allow up to sixteen quarters of revision for original current price and chain volume estimates; that is, the quarters of the current year plus the quarters of the previous three financial years. In these issues, updated annual Supply and Use benchmarks will be introduced.

24.126    In other quarterly releases, revisions to original current price (and chain volume) estimates are restricted to the current year and the previous financial year.

24.127    For chain volume and price measures, the annual re-referencing of the series each September quarter will cause revisions to the levels for the entire series. Re-referencing does not affect percentage movements, but the introduction of updated price weights for the most recent periods could affect growth rates for these periods.

24.128    Revisions resulting from seasonal re-analysis are allowed to flow through to the whole seasonally adjusted time series at the time the seasonal re-analysis is undertaken. This occurs each quarter for series using concurrent seasonal adjustment, and annually for series using annually updated seasonal factors.

24.129    From time to time, the ABS will also implement an 'historical revision' outside of the normal cycle, whereby the whole historical time series is opened up for revision. This allows substantial updates to concepts, classifications, methods and data sources to be reflected across all time periods to which they apply. Since the adoption of the 2008 SNA in 2008-09, significant historical revisions have been made in 2010-11 and 2016-17.

24.130    At present, the following indicative information is provided:

• Australian National Accounts: National Income, Expenditure and Product – Tables detail revisions to the recent eight quarters for GDP and its expenditure, income and industry components. Additionally, the publication contains commentary in the 'Analysis and comments' section on major sources of revisions, and the ‘Explanatory notes’ provide some discussion of the sources of revisions and the accuracy of the estimates more generally.
• Australian System of National Accounts – The major sources of revisions are documented within the 'Analysis of results' section.
• Occasionally, more detailed studies have been made available.

24.131    Analysis of the revisions to the national accounts is undertaken regularly, and the results are made available to users. This is in keeping with developing the Bureau’s broader thinking about data quality declarations for statistical outputs.