Transcript Why is GDP revised` presented by Graeme Walker at the Royal

Why is GDP revised?
Graeme Walker
Head of National Accounts
Royal Statistical Society: 30 May 2012
Outline
•
•
•
•
•
Revisions in general
How GDP is compiled
GDP revisions
Historical perspective
Some numbers
• Levels
• Growths
• Conclusions
• What’s next?
Revisions to economic series
• Revisions are a fact of life for most
economic series
• Or else:
• The first estimate is delayed
• Later information is ignored even when
it tells a different story
• Balance between accuracy and
timeliness
How GDP is compiled (1)
• Three measures
• Output, Expenditure and Income
• Perfect statistical world – all equal
• Many sources
• Monthly, Quarterly, Annual
• Various timings
• Monthly turnover (IoP around 6 weeks)
• Annual tax returns (15 months)
How GDP is compiled (2)
• Preliminary estimate (25 days)
• Accuracy and timeliness trade-off
• One of the fastest in the world
• Based exclusively on output, 44% data
• Second estimate (8 weeks)
• Growth based on output, 83% data
• Publish all measures
• Some provisional expenditure sources
• Limited direct income information
• Use of quarterly alignment adjustments
How GDP is compiled (3)
• Quarterly National Accounts
• 13 weeks
• Based on output, 92% data
• Similar to second estimate
• Expenditure components more firm
• Revisions to previous periods but not for
SU balanced years
• When 4 quarters are published
• Seasonally adjusted annual constrained
to equal unadjusted annual
How GDP is compiled (4)
• Balanced supply and use
• Framework for confronting differences
• Levels not growths
• 110 industries and products
• First balance after around 18 months
• BB12 balances 2010
• Not all benchmarks always received
• Second balance is usually “final”
except for methodological changes
Why is GDP revised?
•
•
•
•
•
•
•
Output source data revised
Expenditure and income
Revised seasonal factors
Annual chain-linking changes weights
Supply and use balancing
New methods
New international frameworks
Historical perspective
•
•
•
•
•
•
Before 1989 – 4 measures published
Output, Expenditure and Income
Average (unweighted)
First estimate (11 weeks)
Poor service sector coverage
Construction
• q1 87 revision
• No attempt to balance the measures
1988 Autumn Statement
• Difficult to assess how economy has been
performing in first half of 1988
Historical perspective
• Pickford review led to current
methodology
• Improved short term surveys
• Balanced measure for years up to t-2
• Chain linking in 2003
• Avoided major revisions with 5 year
rebasing
Some numbers: Revisions to GDP
levels
• Balancing supply and use fixes
annual level not quarterly growth
• Use quarterly path from output
anchored to SUT levels
• Generally small apart from methods
changes
1991 Q1
1991 Q4
1992 Q3
1993 Q2
1994 Q1
1994 Q4
1995 Q3
1996 Q2
1997 Q1
1997 Q4
1998 Q3
1999 Q2
2000 Q1
2000 Q4
2001 Q3
2002 Q2
2003 Q1
2003 Q4
2004 Q3
2005 Q2
2006 Q1
2006 Q4
2007 Q3
2008 Q2
2009 Q1
2009 Q4
2010 Q3
2011 Q2
CVM GDP at Basic Prices, as at BB 2011
380000
360000
£ millions, 2008
340000
320000
BB 2003
300000
BB 2004
BB 2005
BB 2006
280000
BB 2007
BB 2008
260000
BB 2009
BB 2010
240000
BB 2011
220000
200000
The Recession: Q1 2008 to Q3 2009
Revisions to Growth
• Preliminary to QNA growth
• Last 20 periods
• Average -0.02 p.p: not biased
• Absolute average 0.13 p.p
• 5 times unchanged;10 times by 0.1 p.p
• 3 times by 0.2 p.p
• Q4 2009 revised by 0.3 p.p
• Q1 2009 revised by -0.5 p.p (Blue Book)
• No benefit in delaying preliminary
estimate from 25 days to 13 weeks
QNA to Latest
• Evidence from the recession incomplete
• No evidence of bias yet
• Actual revisions now negative
• +0.2 p.p for period 1998 to 2007
• -0.2 p.p for period 2008 and 2009
• Absolute average revisions higher
• 0.4 p.p for period 1998 to 2007
• 0.7 p.p for period 2008 and 2009
• Impact of methodological changes
• Continue to monitor closely
Case study: Q3 2009
• Published in Oct 09 as a fall of 0.4%
• Revised to -0.3% in Nov 09
• Services sector output
• -0.2% in Dec 09
• Quarterly construction data
• Back to -0.3% in Feb 10
• Finance and Business services
• +0.2% in BB 2011, revision of 0.5 p.p
• 0.2 p.p from new sources and SUT
• 0.3 p.p from changing deflation methods
A reminder
Conclusions
• Revisions between Preliminary and QNA
continue to be small and unbiased
• Subsequent revisions have been higher since
2008 but still unbiased
• Partly explained by BB11 methods changes
• Impact of Supply and Use
• ONS monitoring methods through the
recession
• BB 2012 will provide more evidence
What’s next?
• BB 2012 – Improvements to the treatment of
insurance in line with international best
practice
• BB 2013 – More methods improvements;
Whole of Government Accounts likely; full
scope to be published later this year
• BB 2014 – New National Accounts framework
(ESA10)
• Analyse revisions by reason rather than over
time (feasibility)
• Continue to monitor revisions
References
Revisions policy
http://www.ons.gov.uk/ons/guide-method/method-quality/specific/economy/revisions/economic-statistics/nationalaccounts-revisions-policy.pdf
Detail of Supply and Use tables structure
http://www.ons.gov.uk/ons/guide-method/method-quality/specific/economy/input-output/input-output-supply-anduse-tables-structure-overview.pdf
Whole of Government Accounts
http://www.ons.gov.uk/ons/rel/psa/public-sector-finances/psf-and-wga/comparison-of-public-sector-financemeasures.pdf
Revisions triangles
http://www.ons.gov.uk/ons/publications/re-reference-tables.html?edition=tcm%3A77-240821
Improvements to the measurement of Insurance Services
http://www.ons.gov.uk/ons/guide-method/method-quality/specific/economy/national-accounts/methodology-andarticles/2011-present/blue-book-2012-insurance-services/index.html
Various National Accounts articles, including "National Accounts: A short guide"
http://www.ons.gov.uk/ons/guide-method/method-quality/specific/economy/national-accounts/methodology-andarticles/2011-present/index.html
Questions
Email: [email protected]
http://www.ons.gov.uk/ons/rel/naa1-rd/national-accounts-articles/why-is-gdp-revised-/index.html