Page - (000353) - in Knowledge and Networks

351 firms that had collaborated at least twice (i.e., excluding the one-shot collaborators). As for the continuity of linkages, we found that 60 of the 293 linkages were persis- tent and that 138 did not recur. Most of the 293 linkages had been repeated once, and the maximum number of times that a link was subsequently observed to have recurred was 6. Variables We aim to explain the reappearance of linkages that were established between 1983 and 2010. Assume, for example, that we observed a certain firm to have cooperated with a partner in 1997 and that this link recurred in 1998. This activity is what we call repeated cooperation. Assume also that recurrence of this link ceased from 1999 on. With our analysis we seek to explain why the variable for cooperation (the Table 16.1 Description of firms in the sample analyzed for the dynamics of cooperation, 1983–2010 Actors Characteristics No. Size of the pool of potential partners 2369 Cooperating firms 197 One-shot 106 Repeaters 91 Hop-on-Hop-off 27 Mixed-type 40 Persistent 24 Partner diversity (collaboration partners of focal firms) Minimum 1 Maximum 17 Median 2 Links Possible links 321,683 Realized links 293 Repeated links 60 Nonrecurring link 138 Continuity of links (distribution of linkages across times of repetition, without duplicates) 0 138 1 41 2 11 3 3 4 3 5 1 6 1 16 Coevolution of Innovative Ties, Proximity, and Competencies