The geography of collaborative knowledge production: entropy techniques and results for the European Union



records containing at least one address located in an EU member state.8 I further reduced the
size of the dataset by excluding records containing the most common inter-institutional
collaborations with the major countries outside the EU.9 The resulting number of records
amounts to over 200.000 on average each year, with the number increasing from 183.020 in
1993 to 230.561 in 2000.

To sample the number of inter-institutional collaborations within and between European
member states, I used only the first three listings of addresses. Each first and second address,
each first and third address, and each second and third address were counted as one
collaboration. Thus a single -address record yields no collaboration, a double -address record
yields at most one collaboration between a pair European countries, and a record containing
three (or more) addresses yields at most three collaborations between pairs of European
countries.10

4.1 Collaboration in the European Union as a whole

Fig. 1 shows the values of mutual information T as computed from formula (3) for each year
during the period 1993-2000. As explained above, a lower
7' value indicates lower levels of
biases in the choice of partners, and thus a higher level of integration between the EU member
states. The trend of
7' values indicates a gradual integration process suggesting hat EU
member states indeed, on average, have become less biased with regard to the country of
origin of their research collaboration partners.

FIGURE 1 AROUND HERE

The integration process, however, is very slow as the integration indicator decreased only
from 1.526 to 1.461. Put another way, the level of integration in 2000 is 95.7 percent of the
level in 1993. This seemingly slow process of European integration, however, should not be

8 Member states are Austria (AU), Belgium (BE), Denmark (DE), Finland (FI), France (FR), Germany
(GE), Greece (GR), Ireland (IR), Italy (IT), Luxembourg (LU), The Netherlands (NE), Portugal (PO),
Spain (SO), Sweden (SW), and United Kingdom (UK). Note that the United Kingdom refers to records
in the
Science Citation Index containing addresses from England, Northern Ireland, Scotland, or Wales.

9 Being collaborations between an EU country and either Canada, Japan, Russia, Switzerland, or United
States.

10 The use of only the first three addresses provides, apart from computational advantages, a way to
circumvent the fact that in some disciplines collaboration is much more common than in others.

11



More intriguing information

1. The name is absent
2. The name is absent
3. Ahorro y crecimiento: alguna evidencia para la economía argentina, 1970-2004
4. Income Mobility of Owners of Small Businesses when Boundaries between Occupations are Vague
5. EMU: some unanswered questions
6. The name is absent
7. The name is absent
8. The name is absent
9. 03-01 "Read My Lips: More New Tax Cuts - The Distributional Impacts of Repealing Dividend Taxation"
10. Nurses' retention and hospital characteristics in New South Wales, CHERE Discussion Paper No 52
11. The name is absent
12. The name is absent
13. The name is absent
14. Labour Market Institutions and the Personal Distribution of Income in the OECD
15. Olive Tree Farming in Jaen: Situation With the New Cap and Comparison With the Province Income Per Capita.
16. AN IMPROVED 2D OPTICAL FLOW SENSOR FOR MOTION SEGMENTATION
17. Models of Cognition: Neurological possibility does not indicate neurological plausibility.
18. Strategic Planning on the Local Level As a Factor of Rural Development in the Republic of Serbia
19. DISCUSSION: ASSESSING STRUCTURAL CHANGE IN THE DEMAND FOR FOOD COMMODITIES
20. The name is absent