TABLE 2
FROM:
Small states, international pressures, and interlocking directorates: the cases of Switzerland and the Netherlands
Eelke M Heemskerk and Gerhard Schnyder
BACK TO ARTICLETable 2. 'Small World' statistics for the network of interlocking directorates
| N | Avg Degree | Avg Geodesic L | Clustering Coefficient | L (Random) | C (Random) | SW Statistic | |
|---|---|---|---|---|---|---|---|
| CH 1990 | 96 | 8.19 | 2.63 | 0.40 | 2.17 | 0.09 | 3.89 |
| CH 2000 | 95 | 4.36 | 3.81 | 0.42 | 3.09 | 0.05 | 7.47 |
| NL 1996 | 194 | 7.98 | 3.02 | 0.38 | 2.54 | 0.04 | 7.83 |
| NL 2001 | 137 | 7.02 | 3.17 | 0.35 | 2.52 | 0.05 | 5.49 |
Source: The figures were computed based on our data sets using UCINET 6.81 (Borgatti et al., 2002).
N=number of companies; Avg Degree=average Freeman degree centrality; Avg Geodesic=average 'longest shortest path' between each pair of connected nodes; Clustering Coefficient=measuring the proportion of a node's neighbors which are connected among each other; L(random)=expected average geodesic in a random graph of similar size and similar number of ties; C(random)=expected clustering coefficient in a random graph of similar size and similar number of ties.
SW statistic: The small world statistic is defined as the ratio of the actual clustering coefficient and the clustering coefficient of a random graph, divided by the ratio of the actual average path length and the average path length of a random graph (Kogut and Walker, 2001). The higher the score, the stronger are the network's small world characteristics.
It should be noted, however, that Newman et al. (2001) find the random graph comparison to be biased when the bipartite structure of a network is ignored. For purposes of comparisons to previous studies, we show the unipartite statistics. The SW statistics does decline for the bipartite graph (results available on request) but still indicates a small world.
The main result from the table is that the Swiss network's small world character has increased during the 1990s, while the Dutch network's small world character has decreased. This shows that the decline in the number of ties in two different cases may have very different implications at the level of the outcome.
