Transcript Slides - World KLEMS

WORLD KLEMS
CONFERENCE
August 2010
Harvard University
KLEMS Methodology
Basic Building Blocks:
Measuring Capital
Paul Schreyer, OECD
This talk
• The origins and basic ideas
• Capital services and the national accounts –
the debate
• Looking ahead in capital measurement
The origins
• Flow of services and user costs discussed
by Léon Walras (1874) and Böhm-Bahwerk
(1891)
• Model of capital as a factor of production
introduced by Dale Jorgenson (1963)
Capital Theory and Investment Behaviour
• Application to growth accounting by
Jorgenson and Griliches (1967) The
Explanation of Productivity Change
• Large body of literature followed
The basic idea – 2 dimensions
to capital
• Source of flows of services into production
(the ‘K’ in the production function) 
Production aspect
– Captured by recognizing that different types
of assets have different marginal
productivities
– These are captured by asset-specific prices of
capital services: user costs
– User cost weights are key in aggregating
across different types of capital (akin to
labour input)
• Component of wealth  Wealth aspect
The basic idea (2)
• Traditionally, only wealth aspect
recognised in NA, via balance sheets
• Production aspect, if recognised at all,
only in a partial and inaccurate way:
depreciation as measure of capital input
into production
• Yet, both aspects can be integrated into a
system of accounts in a conceptually
correct and consistent way
Balance sheet
Age -price
function
Net capital
stock
Accumulation
accounts
Net value added
Production
account
Return on capital
CFC
User costs
Gross
stock
Investment
Generation of
income
account
Retirement
function
Age -efficiency
function
Productive
stock
Capital services
The basic idea (3)
• Consequence:
• Symmetric treatment of labour and capital
in the production (and generation of
income) accounts
• Both the value of labour income and the
value of capital income can be broken
down into a price and volume component
 vital for productivity computations
• SNA 59, SNA 68 and SNA 93 had
asymmetric treatment of labour and
capital
Capital services and the
national accounts – the debate
• First discussions in the context of the
‘Canberra I Group on Capital
Measurement’ but notion of capital
services was not adopted
• More discussions in Canberra II Group
(2005-2008) and recognition of capital as
an input in production measure put
forward as issue in the Revision of SNA 93
 SNA 2008
• 5 issues in the debate
Issue #1: “Capital services
measures mix up the real and
the financial side of the
economy”
• Misunderstanding about the role of rates
of return and holding gains and losses in
user costs
• Absent market transactions, they form
part of an estimate of an imputed
transaction but not a recognition of capital
gains as production
Issue #2: “Government
production is not for profit –
hence no net return on capital
should be imputed for
government assets ”
• Neglects the fact that government has financing
costs or opportunity costs
• But SNA 2008 maintains that value of capital
services for government asset = depreciation
• Ironically, net return is needed to identify nonmarket producers
Issue #3: “There is no unique
method to estimate user costs –
this will impair international
comparability”
• No misunderstanding here but there are
many instances in national accounting
when different methods can be used and
are used
• Real questions:
– How much does it matter (N.Oulton’s paper)
– Is not recognising user costs a better option?
Issue #4: “Already too many
imputations in the accounts – no
need to add another one”
• Fair point
• Judgement about optimal level of
imputations varies between national
accountants and analysts
Issue #5: “Capital service
measures imply a value
judgement by the statistician on
functional income distribution”
• Capital services measures are indeed
conceived with regard to a particular
theoretical model (neoclassical theory)
• But issue is to be transparent about this.
There is (and probably should be) no such
thing as a theory-free national accounting
Do capital services measures
evolve differently from capital
stocks? Yes – example Australia
280.0
260.0
240.0
220.0
200.0
180.0
160.0
140.0
120.0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
100.0
Capital stock
Capital services
Looking ahead in capital
measurement
• Linking industry and total economy
accounts
• Capitalisation of intellectual property
products beyond R&D: the debate on
intangibles
• Capacity and capital utilisation
• Natural and environmental assets
• Human capital
For more, and for references to
the literature see...
Also available in French, Spanish and ...Korean
Thank you!
[email protected]
OECD Productivity data
www.oecd.org/statistics/productivity/