Key Figures 2005 on Science, Technology and Innovation

1
Key Figures 2005 on Science, Technology and
Innovation The Composite Indicators on the
Knowledge-based economy
Vincent Duchêne, DG RTD M2 WORKSHOP on EUROPEAN
COMPOSITE INDICATORS  SCOREBOARDS Brussels,
24-25 October 2005
2
Key Figures on STI

• Published yearly since 1998
• Main indicators (OECD, Eurostat) to take stock of
  position of EU (and Member States) in Science,
  Technology the Knowledge-Based Economy,
  vis-à-vis main competitors (based on set of /-
  35 indicators)
• Usually consists of 3 parts
• Performance of the EU in the global
  macro-economic context
• Economic growth, Employment, Productivity
• 2 Composite indicators on the KBE (Investment in
  the KBE, Performance in the transition towards
  the KBE)
• Input-side - Investment in the KBE
• RD Expenditure, Execution of RD, patterns of
  financing RD (links private RD public RD
• RD spending by level of technology, by size
  class
• Complementary sources for financing of domestic
  RD International flows of RD spending,
  High-tech Venture capital, tax incentives to RD
• Education expenditure on (higher) education,
  Human Resources in ST (researchers, PhDs,
  international migration)

3
Key Figures on STI

• 3. Output-side - Performance of the KBE
• ? Scientific output scientific publications,
  citations, co-publications
• ? Technological output patenting activity,
  co-patenting, ICT- Biotech-patents
• ? Competitiveness of High-tech industries
  high-tech trade, sector-specific productivity
  performance
• Innovation of 2005 edition
• Data up to 2003 ? First time official data one
  year after Barcelona Council (March 2002) 3 -
  2/3 targets
• Integration China

4
CIs on the KBE Why ?

• Since 2002 edition of KF, integration of 2
  Composite Indicators on the knowledge-based
  economy
• CI of Investment in the KBE
• CI of Performance in the transition towards the
  KBE
• Requested in July 2001 by Commissioner for
  Research Ph. Busquin
• Context Lisbon strategy structural indicators
• Aim
• Single measure (indicator) to capture progress
  towards KBE (knowledge dimension of Lisbon
  strategy)
• Attract media attention promote support for KBE
  related policies
• Reinforce / broaden monitoring of section
  Innovation Research in Structural Indicators
  database

5
CIs on the KBE Process

• Pre-conditions
• Transparent and easy-to-understand
• Clear political message
• Broad base (KBE more than technology)
• Robust Methodology
• Iterative, collective process
• Co-operation with EC Services (Estat, EAC, Infso,
  Entr, JRC-Ispra, Ecfin), national stat offices
  (through ESS), OECD
• External assessment by MERIT() on
  appropriatedness of
• Methodologies
• Constituent indicators
• () ARUNDEL, A. and BORDOY, C., Methodological
  Evaluation of DG Researchs Composite Indicators
  for the Knowledge-based Economy, MERIT, June 2002)

6
CIs on the KBE 4 steps

• Selection appropriate indicators
• Normalisation / standardisation measurement
  units
• Weighting scheme
• Robustness tests (checking steps 1 to 3
  imputation missing data cfr. Presentation
  Stefano Tarantola)
• Two Composite Indicators on the KBE
• Investment in the KBE (capacity to create /
  absorb new knowledge)
• Performance in the transition towards a KBE
  (capacity to transform new knowledge tech.
  econ. progress)

7
1. Selection Component indicators

• 2 dimensions in Investment in the KBE
  creation and diffusion of knowledge
• 4 criteria for selection
• Policy relevance
• Reliability
• Coverage (over time EU-15)
• Redundancy (Principal Component Analysis)

8
2. Standardisation Method

• 2 selection Criteria
• simple and easy-to-understand
• low sensitivity to outliers
• Selection
• M1 ok, but no interval-data
• M2 more sensitive to outliers
• M3gtM4 M4 less sensitive but M3 simpler

Notes is the value of indicator j for country i
at time t. is the weight given to indicator j in
the composite index. In equation 1, p an
arbitrarily chosen threshold above and below the
mean. Source Arundel and Bordoy, p. 7
9
2. Standardisation Methods 3 4 very similar
results
R²0.998
Source ARUNDEL and BORDOY, 2002, p. 11
10
3. Weightings

• Problem there is no magic formula giving the
  contribution of RD, Education etc to KBE !
• Option 1 Using Factor Analysis (Principal
  Component Analysis) to determine weights and to
  aggregate component indicators !
• Advantages 1) no redundancy anymore, 2)
  weightsloadings of Principal Components
• Disadvantages 1) complexity ! (whats in PC1) ?,
  2) not comparable over time (weights will change
  from year to year since data/indicators change)
• Option 2 Equal weights to each conceptual group
  of indicators (creation vs. diffusion) equal
  weights to each indicator within each conceptual
  group
• Option 2 selected !

11
Component indicators weightings CI on
investment in the KBE
12/24
12/24
12
Component indicators weightings CI on
performance in the transition towards the KBE
13
Composite indicator on Investment in the KBE
1997-2002
14
Country-specific deviations scores where is the
difference ?
15
Investment vs Performance in the KBE Is more
better ?
16
Towards the Second Generation ...

• Integration new / better component indicators
  (ICT, technology diffusion etc)
• Methodological improvements (BoD-method ?
  Imputation missing values ?)
• New Composite Indicators on Framework conditions
  (state aid, public procurement, fiscal incentives
  and other indirect measures, etc) ?