Table 1.
Basic statistics of the largest connected components of boards projections.
Figure 1.
Board interlocks among EU countries in 2005 (Left) and 2010 (Right).
The size of a node is proportional to the square root of the number of companies from that country, the thickness of an edge is proportional to the number of interlocks between the boards of companies from the connecting countries. Each node reports in a pie-chart the distribution of companies among the sectors listed on the right.
Figure 2.
Board interlocks among sectors in 2005 (Left) and 2010 (Right).
The size of a node is proportional to the square root of the number of companies in that sector, the thickness of an edge is proportional to the number of interlocks it represents.
Figure 3.
Assortative mixing in the boards projection in 2005 (Left) and 2010 (Right).
The plots show the behavior of the weighted average degree of a node's nearest neighbors as a function of the node degree. Thin red lines display linear regressions: slope coefficients are and
for 2005 and 2010, respectively, whereas the regression
is
and
.
Figure 4.
Centrality measures on the boards projection in 2005 (Left) and 2010 (Right).
Each point in the plane corresponds to a company; and
coordinates define its betweenness (here renamed brokerage), and eigenvector centrality (here called accessibility) scores, respectively. Due to readability reasons, only those who rank higher are labeled (refer to Tables 2 and 3).
Table 2.
Top 10 ranked firms with respect to three different measures of centrality – Year 2005.
Table 3.
Top 10 ranked firms with respect to three different measures of centrality – Year 2010.
Figure 5.
Brokerage against transnational character in 2005 (Left) and 2010 (Right).
Each point in the plane corresponds to a company; and
coordinates define its normalized node betweenness (brokerage) and its ratio of transnational edges, respectively. For readability reasons, only company names in the upper
-quantile of the betweenness distribution are shown.
Figure 6.
Communities in the largest connected components of firms projection in 2005 (Left) and 2010 (Right).
The size of a node is proportional to the square root of its number of connections, the thickness of an edge is proportional to the number of shared directors between the connecting boards. Colors refer to the best-found community partition, whose weighted modularity score is for 2005 and
for 2010.
Table 4.
Country distribution of the best-found subdivision into communities – Year 2005.
Table 5.
Country distribution of the best-found subdivision into communities – Year 2010.
Figure 7.
Communities in the subgraph of firms projection in 2005 (Left) and 2010 (Right) considering transnational links only.
The size of a node is proportional to the square root of its number of connections, the thickness of an edge is proportional to the number of shared directors between the connecting boards. Colors refer to the best-found community partition, whose weighted modularity score is for 2005 and
for 2010. Note that the largest connected component covers
of the transnational subgraph in 2005, and
in 2010.