Chain volume measure

Note: This page is copied basically word-for-word from http://www.aph.gov.au/library/pubs/mesi/features/chain.htm. Since this is a publication of the Australian government, this copying is less obviously illegal than it might otherwise be?

Chain Volume Measure (CVM) is an index number .

In publications of statistical data, for example Gross Domestic Product (GDP), there may be a necessity to compare data at different time periods. The choice of a year on which to base such comparisons is important. Fixed base aggregations compare data in one period with that of another, for example, data for GDP growth for 2006 might be compared with a base year of 2001. Using this method presents problems because of the way in which the information on which the base year is centred often changes. Video cassette recorders might have been a feature of data in 2000 but not in 2006, for example, and so a fixed based aggregate might not be as accurate as a means of comparison. It is for this reason that chained volume measures (CVM) have been introduced by the Office for National Statistics (ONS) in the United Kingdom. The use of CVM allows the base year to be updated more regularly and thus the statistics will be more accurate. Using this method the base year is revised every year rather than every five years that was the case. Calculations are carried out in previous years prices and aggregated to give chain volume measures. The ONS has replaced the term 'constant prices' in its terminology with CVM to describe components of GDP.

Chain Volume Measures

The release of the September quarter 1998 national accounts by the Australian Bureau of Statistics (ABS) sees changes associated with the implementation of new international standards contained in the System of National Accounts 1993 (SNA93)(1) Two of the main impacts are:

• a single measure of gross domestic product (GDP). The three presently published measures-based on the income, expenditure and production approaches-will be discontinued, as will their average [GDP(A)], and
• constant price estimates will be replaced by chain volume measures, which SNA93 argues will provide better indicators of volume (i.e. real) growth.

This feature article will confine itself to a discussion of the replacement of the constant price estimates.

Constant Price Estimates

For many years real estimates of expenditure and production (output) have been provided in constant price form, i.e. where the effects of changes in prices have been removed.

The current value of each transaction recorded in the national accounts can be thought of as a quantity multiplied by a price. Therefore to produce a constant price estimate the current price is replaced with the corresponding price in a base year. Up to the last release of the Australian national accounts this was the financial year 1989-90.

Although these constant price estimates were expressed in dollar terms they only change with changes in the underlying quantities because the prices have been held constant.

Criticism of Constant Price Estimates

The major criticism of the constant price estimation process is that the estimates usually become less accurate the further the current period is from the base year. This is particularly so when the price relativities (i.e. the ratios of the prices of commodities to the prices of other commodities) of goods and services change substantially.

There are various methods that can be used to derive constant price estimates, each producing different results depending on the extent to which relative prices change. If all prices increase at the same rate then the methods of producing constant price estimates will produce the same results.

However, in reality, this is not the case. Some prices increase faster than others and some prices fall while others rise. The more variation in relative price changes the greater the difference between constant price estimates that can be produced and the further these estimates are from a 'true' real value.

This is known as the substitution effect. Items of rapidly increasing price are replaced (substituted) by items that have become relatively cheaper. For example some items, such as computing equipment, have had significant price falls, and subsequent increased quantities, in recent years and have been attributed too much weight in constant price estimates; this has caused an overstatement of growth.

The major question, then, is how often should a constant price series be rebased to minimise this substitution effect. Published constant price estimates have usually been rebased about every five years.

Chain Volume Measures

SNA93 recommends that this rebasing should be done annually and the ABS will now follow some of the world's leading statistical agencies by adopting this recommendation. As with constant price estimates, chain volume measures will be expressed in dollar terms.

By rebasing annually, chain volume measures will always provide price relativities that reflect the current situation. This will provide better real estimates, especially in times of rapidly changing relative prices.

Chain volume measures are derived by linking together (compounding) movements in volumes, calculated using the average prices of the previous financial year, and applying the compounded movements to the current price estimates of the reference year.(2)

This annual rebasing will mean that each year the chain volume measures will have a different base year. The base year will always be the previous complete financial year. With the release of the September quarter 1998 data the base year will be 1996-97. The release of June quarter 1999 figures will have a base year of 1997-98 and the base year will change each year with the release of the June quarter data.

When a new base year is introduced the level of the chain volume measures series will change but the growth rates will remain unaltered unless caused by other factors.

Where a constant price series has been used in the past the corresponding chain volume measure series should now be used.