National Statistics Online - Articles - The effect of annual chain-linking on the Blue Book 2002 annual growth estimates
Articles Database
The effect of annual chain-linking on the Blue Book 2002 annual growth estimates
A new method for estimating economic growth would have had a maximum impact of +/-0.2% on economic growth as published in Blue Book 2002.
This article has been amended since its publication on 16 April 2003. The following corrections have been made:
In the first paragraph of the conclusion, the original penultimate sentence 'There are small upward effects from 1999 to 2001' now reads 'There are small downward effects from 1999 to 2001'.
In the final sentence of the article, the original, 'the upward effect on GDP growth rates is greater' now reads 'the effect on GDP growth rates is greater'.
Chained volume measures for GDP and GDP components are due to be published for the first time in the UK on September 30th 2003. This article describes the results from test-running the new annual chain-linking (ACL) systems used for compiling these chained volume measures using inputs consistent with Blue Book 2002. The results show that the effects on GDP growth rates published in Blue Book 2002 as a result of ACL are no larger than +/-0.2 per cent from 1995 to 2001.
These effects are in line with those predicted by economic theory. Analysis of the results shows that the small net negative effect of ACL on recent annual growth can be attributed to expenditure on computers, investment in machinery and equipment and trade in high tech goods in the expenditure measure of GDP.
A similar small net negative effect on the output measure of GDP can be attributed to the production of communication services partially counterbalanced by a positive contribution from manufacturing. This kind of comparison presents the only opportunity to isolate the effects of ACL on estimates before revisions to data inputs are introduced as part of the usual Blue Book process which will interact with ACL and the sources of these other revisions are listed in this article. Final revisions to growth will therefore differ from the effects shown here.