Extending the asset boundary to include Research and Development Issues 9 and 10

1
Extending the asset boundary to includeResearch
and DevelopmentIssues 9 and 10

• Charles Aspden
• OECD

2
Presentation

• Canberra II and NESTI Groups
• Current situation in the 1993 SNA
• Why is the way it is?
• Obstacles to change
• Issues addressed by Canberra II
• Next steps
• Recommendations

3
Canberra II and NESTI Groups

• Canberra II Group has discussed this issue at
  every one of its five meetings
• Close cooperation with the NESTI Group
• particularly Soli Peleg and Pablo Mandler from
  Israel.
• Comprehensive consideration of conceptual and
  practical issues
• Further cooperation with NESTI if proposals
  accepted

4
Current situation in the 1993 SNA

• Expenditures on RD not GCFC, despite recognition
  that they are inherently investment in nature.
• Instead, SNA recognises patented entities,
  which are non-produced and appear in the other
  changes in assets account.

5
Current situation in the 1993 SNA

• In effect, the SNA
• recognises those assets created by RD which
  generate an income flow between units
• but denies their connection with RD production,
  and
• does not recognise RD assets whose services are
  consumed by their owners

6
Current situation in the 1993 SNA

• 1993 SNA differs from 1968 SNA in that income
  flows associated with patented entities are
  recorded as income from sales and use of a
  service, rather than property income.
• Clearly, the authors of the 1993 SNA thought that
  patented entities were really produced assets.

7
Why is the 1993 SNA the way it is?Explanation in
para 6.163

• Need to have clear criteria for delineating RD
  from other activities.
• Be able to identify and classify the assets
  produced.
• Be able to value the assets and depreciate them.

8
Issues addressed by Canberra II Group

• Scope of RD GFCF
• Definition of RD
• Measurement of GFCF
• Avoidance of double counting
• Constant price estimation
• Measurement of capital stock, capital services
  and consumption of fixed capital

9
Scope of RD GFCFShould all of RD be included?

• Nearly all Canberra II members think expenditures
  on RD should be recorded as GFCF.
• The majority favour including all of it either
  for conceptual reasons, or because the parts they
  think should be excluded (esoteric RD) are so
  small that it is better not to try to exclude
  them.
• A minority would like to exclude RD output made
  freely available.

10
Scope of RD GFCFShould all of RD be included?

• Reasons for including all RD
• RD is thought to be one of the principal drivers
  of output growth. It is important for
  productivity analysis across time and across
  countries that RD data be fully integrated in
  the national accounts.
• From an economic theory point of view, all RD
  expenditure qualifies as capital formation
  because consumption is foregone for the sake of
  future benefits.

11
Scope of RD GFCFShould all of RD be included?

• It is important to include all forms of RD,
  including basic research - some of which may be
  undertaken without any specific goal in mind.
  Without such research technological progress
  would be restricted to improving existing
  technologies and technological leaps would occur
  by chance. Clearly, those who finance basic
  research do so in the belief that their will be
  future benefits. If we exclude it, we could be
  excluding an important factor influencing
  economic performance.

12
Scope of RD GFCFShould all of RD be included?

• Units that make their RD output freely available
  often undertake the RD with the expectation of
  gaining future benefits, e.g.
• Medical research to provide more cost-effective
  treatments
• Studies of genetics to identify people at risk of
  illness or prone to criminality.

13
Scope of RD GFCFShould all of RD be included?

• Both private and public sectors make some RD
  output freely available and protect the rest.
  Although the proportions differ.
• Government institutions, such as universities
  often take out patents.
• Defence RD is kept secret.

14
Scope of RD GFCFShould all of RD be included?

• In practice it would be very difficult to exclude
  RD made freely available accurately, and
  attempting to do so could add substantially to
  the estimation costs. One could expect reduced
  comparability over time and across countries.
• It would be very crude to try to exclude RD made
  freely available by excluding all general
  government RD.

15
Scope of RD GFCFShould all of RD be included?

• The SNA recommends that non-market output should
  not be valued by the amount of income received.
  This extends to the freely available services of
  public roads, whose asset values are determined
  on the basis of cost.
• Canberra II believes this recommendation should
  include the services of RD assets made freely
  available.
• The definition of an asset should be clarified to
  cover such assets.

16
Measurement of GFCF

• As most RD is conducted on own account, it is
  proposed to use expenditure data collected as per
  the Frascati Manual as the principal data source.
• 89 countries have RD data on the UNESCO
  database, mostly time series.
• Request submitted to NESTI for changes to FM to
  better accommodate SNA needs.

17
Definition of RD

• The FM defines research and experimental
  development to comprise creative work undertaken
  on a systematic basis in order to increase the
  stock of knowledge, including knowledge of man,
  culture and society, and use of this stock of
  knowledge to devise new applications.

18
Definition of RD

• The 1993 SNA does not give an explicit definition
  of RD and the closest it comes to doing so is in
  paragraph 6.163 research and development are
  undertaken with the objective of improving
  efficiency or productivity or deriving other
  future benefits so that they are inherently
  investment rather than consumption type
  activities.
• However, various paragraphs in the 1993 SNA that
  mention RD imply a definition that is somewhat
  broader and more akin to the FMs definition and
  coverage of market and non-market activities,
  including, for example, RD activities that are
  mixed with education (paragraph 9.89).

19
Other issues

• Avoidance of double counting solutions have been
  found to major problems.
• Constant price estimation research is being
  undertaken to develop output price indexes for
  RD, but well-constructed input price indexes
  will suffice.
• Capital stock, etc. Much work has been done to
  estimate asset lives and depreciation rates, but
  more work needed.

20
Practicality of implementation

• A number of Canberra II members have compiled RD
  accounts in order to determine the practicality
  of doing so and to identify problem areas.
• This work has shown that it is feasible to
  compile RD estimates in the national accounts of
  adequate quality.
• Planned future work is aimed at improving the
  quality.

21
Next steps if recommendations accepted

• Canberra II to work with NESTI to change or
  augment RD surveys to better meet SNA needs.
  First meeting tentatively set for early 2006.
  Also need to work on improving the measurement of
  international trade in RD.
• Canberra II Group to undertake further research
  on asset lives and depreciation rates for RD.
• NESTI Group working on the development of output
  price indexes for RD.

22
Recommendations

• The 1993 SNA should be changed to recognise the
  outputs of RD as assets, and the acquisition,
  disposal and depreciation of RD fixed assets
  should be treated in the same way as other fixed
  assets.
• All RD output should be treated as an asset,
  irrespective of its nature or whether it is made
  freely available. In the latter case, the asset
  should be recorded on the balance sheet of the
  owner of the original and be regarded as
  providing a free service until it becomes
  obsolete.

23
Recommendations

• The definition of an asset should be reviewed to
  ensure it covers the assets of non-market
  producers adequately.
• The definition of RD given in the Frascati
  Manual (FM) should be adopted in the SNA.

24
Recommendations

• The Frascati system provides the best means of
  deriving estimates of RD statistics, principally
  gross fixed capital formation (GFCF). However,
  there are shortcomings in the Frascati data and
  the FM should be amended to better support the
  needs of the SNA. (NESTI has indicated a
  willingness to do this.)
• Most RD output is produced over several periods
  and the SNA recommendations for the production of
  other assets should apply. Most RD production is
  on own account, which implies recording it as
  GFCF as it occurs under the current
  recommendations.

25
Recommendations

• Patented entities should no longer be recognised
  as assets in the system.