1 of 25

Understanding Innovation Indices: Statistical Properties and Macroeconomic Linkages

Scott W. Hegerty, Ph.D.

Distinguished Professor of Economics, NEIU

World Economy Research Institute

October 23-24, 2025

​

2 of 25

Main idea: �Examine the Global Innovation Index

  • Some questions:
  • How does the GII behave over time?�🡪 in CEE/Poland?
  • Where is it most/least stable?
  • Which countries are ranked highest/lowest?
  • Is it different from merely knowledge? Related to other indices?
  • Is it related to growth?

3 of 25

Main idea: �Methods

  • Construct annual time series for as many countries as possible
  • Examine stability in levels and ranks
  • Look at relevant subcomponents�🡪 Comparisons to GII?
  • Compare to other measures (such as SII)
  • Growth correlations

​

​

​

4 of 25

Data

  • Global Innovation Index: From WIPO website �(separate spreadsheets)
  • Files combined; only ones with all years used
  • Subindices: Examine and compare

​

  • Correlations:
  • Macro variables (growth)
  • Summary Innovation Index

5 of 25

GII Index:

  • 217 countries (2013-2022), some blanks🡪 � 116 with all years (2013-2024)
  • Subcomponents: 154 total�(Not all complete)�(An example: "Global Innovation Index: Madrid system trademark applications by country of origin" )
  • Focus on Knowledge Creation, Knowledge Diffusion, Knowledge Absorption

​

​

6 of 25

Subcomponents

  • Some examples:

​

7 of 25

Subcomponents as an alternative measure

  • Knowledge Creation by itself (is this similar to the index?)
  • Three Knowledge variables combined as new index

​

  • Principal Components Analysis:�🡪 New index from three components�🡪 Linear combination that maximizes variance�(Basic results: Eigenvalues > 1 , factor loadings)

​

8 of 25

Statistical Measures

  • Ranks, levels (scores), variation, stability
  • Mean, SD: Summary statistics
  • Coefficient of Variation (SD/mean)

​

  • Focus on CEE and Poland

​

9 of 25

Correlations

  • Summary Innovation Index (SII), 46 countries� European Innovation Scoreboard

​

  • GDP growth, all 116 countries� World Bank data

10 of 25

Results: Principal Components

​

  • Indices generally not correlated with GII
  • CEE countries evaluated specifically

​

11 of 25

Correlations:� GII and PCA

Results: PCA vs. GII, CEE

Example: Poland’s eigenvalues/loadings

12 of 25

Green = CEE Red = Poland; means over all years

Results: Principal Components vs. GII

13 of 25

Mean vs. CV

​

  • Larger means -> lower variation
  • Innovation leads to stability?

​

Poland: Right in the middle (GII) but low mean/high variance for PCA

14 of 25

Green = CEE Red = Poland

Y = 23.04 -0.379X, R2 = 0.591

​

15 of 25

Mean vs. CV (Knowledge Creation)

Green = CEE Red = Poland

16 of 25

Mean vs. CV (PCA)

Green = CEE Red = Poland

17 of 25

Stability in Scores

  • Movement is not that large in the grand scheme of things
  • But can still isolate some big drops or increases
  • Using ranks also can help quantify this

18 of 25

Values (zoomed in)

19 of 25

Values (zoomed out)

20 of 25

Ranks (low = high innovation)

Mean = 41

21 of 25

Biggest changes in ranks

Deteriorations vs. improvements

22 of 25

(no) Relationship between rank change and average GDP growth

Spearman correlation = -0.047

23 of 25

Main findings

  • Interesting descriptive analysis:�Index components�Which countries have highest/lowest overall ranks�Changes over time
  • Mean/variation relationship (innovation -> Stability)
  • Relative position of CEE and Poland (in the middle of the pack_
  • Lack of Correlations:�Among GII, PCA, SII�Between GII and growth�

​

24 of 25

Conclusions

  • Innovation and stability are related
  • The GII is a unique measure that is difficult to replicate
  • Poland and CEE are in the “middle” for various criteria
  • Little connections between innovation growth�

Next steps:�Case studies: Leaders/laggards (worldwide), CEE

Formal analysis of macro linkages

​

25 of 25

Thank you!

  • S-Hegerty@neiu.edu