Distances between pairs of car manufacturers

The distance between car manufacturers has been calculated based on Dijkstra's (1960) algorithm. This algorithm sums the ‘cost’ of connections and find the path of least resistance, considering the weights of the ties. A strong tie has a low cost, i.e. short distances, whereas a weak tie has a high cost, i.e. long distances.

Newman (2001) has developed a method to apply Dijkstra’s algorithm to a network by taking a tie and defining it as ‘1’ and then dividing 1 by the weight. In our thesis, this is applied by dividing 1 by the number of umbrella memberships the ties represent. For example, BMW and Toyota have a strong tie between them since they both share memberships of seven of the same umbrella organizations. This implies that their tie is equal to 1/7, hence being strong since the strongest tie in the network is made of eight co-memberships as it is stated in the adjency matrix in Table 8. The average geodesic distance in the network has been calculated to 0.39, also using

Newman’s (2001) approach. It was expected that the average distance would decrease when transforming into a one-mode network, since a large set of nodes have been removed (the umbrella organizations), however an average distance of 0.39 presents a network where information can travel very fast and efficient. The one-mode network therefore presents a higher degree of connectedness (which was no surprise since the least connected nodes, i.e. umbrella organizations, have been incorporated in ties instead of representing nodes.

The distances between pairs of actors, are presented in Table 9:

BMW Honda Nissan Mazda Mitsubishi Renault Skoda Suzuki Toyota VW BMW 0.1667 0.2000 0.4500 0.6250 0.1250 0.1667 0.5000 0.1250 0.1667 Honda 0.2500 0.5833 0.3333 0.2917 0.2500 0.3333 0.1429 0.2500 Nissan 0.2500 0.5833 0.2500 0.3333 0.5833 0.1667 0.3333 Mazda 0.8333 0.5000 0.5833 0.5000 0.3333 0.5833 Mitsubishi 0.7679 0.5000 0.6667 0.4762 0.5000

Renault 0.3333 0.6250 0.2500 0.3333

Skoda 0.6667 0.1667 0.1250

Suzuki 0.4762 0.6667

Toyota 0.1667

VW

Table 9: Geodesic distances between pairs of car manufacturers in the network

In the network, the distances between pairs of actors vary within an interval of 0.125 to 0.667. With 0.125 being the lowest distance in the network and representing a relatively strong tie and 0.667 being the largest distance and thereby representing a ‘weaker’ tie. This is a relatively small scale to evaluate ties upon why it must be fair to assume that the network as a whole provides fertile soil for social contagion due to the generally small distances between actors in the network and thereby also strong ties. The average geodesic distance of 0.39 further supports this.

However, some pairs of actors are more likely become similar based on the contagion-argument than others.

The average distance of the individual actors and the distances between pairs of actors are useful measures to determine which actors that are most likely to become similar. Table 10 presents the average distance of the individual actor’s ties, and Table 11 presents the distances between pairs of actors in the network.

Car manufacturer Average geodesic distance

Toyota 0.26

BMW 0.28

Honda 0.29

Nissan 0.33

VW 0.35

Skoda 0.35

Renault 0.39

Mazda 0.51

Suzuki 0.56

Mitsubishi 0.59

Table 10: The average geodesic distance of the individual actor’s ties. A low average distance will imply generally strong ties of an actor. A high average distance will imply generally weak ties of an actor.

Based on the theory of social contagion, this thesis would expect actors having a ‘low’ average distance, hence generally strong ties, to be similar to their relations in terms of political position. On the other hand, this thesis would expect actors having a ‘high’ average distance, hence generally ‘weaker’ ties, to be less similar to their relations in terms of political position. Based on the figures from Table 10, actors having a low average distance include Toyota (0.26), BMW (0.28), Honda (0.29) and Nissan (0.33), whereas actors having a higher average distance include Mazda (0.51), Suzuki (0.56) and Mitsubishi (0.59).

Table 11 ranks pairs of actors based on their distances, i.e. their strength of relations. On the top of the list will pairs of actors with the shortest distance appear, hence pairs of actors most likely to be similar based on contagion-argument. At the bottom of the list will pairs of actors with the largest distance appear, hence pairs of actors that are less likely to be similar based on the contagion-argument (at least compared to the actors above them in the table).

Table 11: Ranking of network pairs based on their distances, i.e. strength of ties. A short distance is equal to a strong tie. A long distance is equal to a weak tie.

Based on the theory of social contagion, this thesis expects that pairs of actors ranked at the top of the list in Table 11 will tend to be similar in their political position. These actors will, due to their strong relational ties, engage in frequent communication and will in situations of uncertainties (i.e. the presence of a possible legislative proposal) tend to discuss the costs and benefits of adapting to the uncertainty. This will create a normative understanding on the structures of the situation, which will make them likely to adopt the same behavior and attitude on the situation. Opposite, this thesis would expect pairs of actors ranked at the bottom of the list in Table 11 to be less likely to have similar political positions due to their ‘weaker’ relational ties.

It should be emphasized from Table 11 that the actors located in the ‘core’ of the network as presented in Figure 7, (expectedly) have short distances to each other, i.e. strong ties. These actors include: BMW, Honda, Renault, Nissan, Skoda, Toyota and Volkswagen. The thesis would therefore expect this small group of actors in the core of the network to be very similar in their political position. Opposite, this thesis would expect actors

Pairs of actors Distance Nissan-VW 0.3333

BMW-Renault 0.125 Mazda-Toyota 0.3333

BMW-Toyota 0.125 Renault-Skoda 0.3333

Skoda-VW 0.125 Renault-VW 0.3333

Honda-Toyota 0.1429 BMW-Mazda 0.45

BMW-Honda 0.1667 Mitsubishi-Toyota 0.4762

BMW-Skoda 0.1667 Suzuki-Toyota 0.4762

BMW-VW 0.1667 BMW-Suzuki 0.5

Nissan-Toyota 0.1667 Mazda-Renault 0.5

Skoda-Toyota 0.1667 Mazda-Suzuki 0.5

Toyota-VW 0.1667 Mitsubishi-Skoda 0.5

BMW-Nissan 0.2 Mitsubishi-VW 0.5

Honda-Nissan 0.25 Honda-Mazda 0.5833

Honda-Skoda 0.25 Nissan-Mitsubishi 0.5833

Honda-VW 0.25 Nissan-Suzuki 0.5833

Nissan-Mazda 0.25 Mazda-Skoda 0.5833

Nissan-Renault 0.25 Mazda-VW 0.5833

Renault-Toyota 0.25 BMW-Mitsubishi 0.625

Honda-Renault 0.2917 Renault-Suzuki 0.625

Honda-Mitsubishi 0.3333 Mitsubishi-Suzuki 0.6667

Honda-Suzuki 0.3333 Skoda-Suzuki 0.6667

Nissan-Skoda 0.3333 Suzuki-VW 0.6667

BMW-Renault 0.125 Mitsubishi-Renault 0.7679

located more in the periphery of this core, i.e. Suzuki, Mazda and Mitsubishi, to be less similar to the group of actors found in the core, since their distances to these actors are longer, hence ties are weaker.