Endogenous Network Effects between Different Types of Interfirm Ties: A Review and Empirical Assessment
Olga NovoselovaThe structure of existing interorganizational networks influences new connections via endogenous network effects. While previous research established such effects between past and future ties of the same type, the growing attention to network multiplexity—various tie types simultaneously connecting organizations—highlights interesting questions about tie-to-tie effects between different connectivity modes. Does the presence, magnitude, and mechanisms of endogenous effects depend on the types of preceding and forming interorganizational connections? Do endogenous effects operate in distinct or unexpected ways in particular pairs of tie types? Are there general patterns of effects, uniform across the multiplex interorganizational network?
Organizational scholars examined interdependencies between different connectivity modes, but it is difficult to derive a systematic picture from such studies due to both the number of possible pairwise effects and their diverse potential mechanisms. While only one effect of past ties on new ones is possible within the same connectivity mode, research on effects between different modes needs to examine all possible pairs of tie types and test the effect of the first on the second and vice versa. Theoretical complexity is also higher: for example, the concept of network inertia can apply to persisting patterns of the alliance network as well as the board interlock network, but it is harder to draw theoretical analogies between the effect of alliance connectivity on board interlocks formation and the effect of connectivity through a common investor on alliance formation.
The first part of this study systematically reviews past research. Based on an expansive search for management and sociology articles addressing the interdependencies between different tie types, it summarizes their findings and classifies previously tested and observed effects. The revealed fragmentary state of knowledge suggests that a comprehensive exploratory study, rather than isolated hypotheses testing, can provide valuable novel information about endogenous network effects and offer promising avenues for future research. The exploratory approach allows to identify those effects that likely exist but have not been studied or hypothesized and to see whether known effects found in particular contexts are observed in the general population of firms. Hence the second part examines the effects of past connectivity on tie formation among the S&P 500 companies in 1997—2015. The analysis includes four classes of effects and four types of interorganizational connections—through an alliance, board interlock, executive migration, and investor interlock—covering sixty endogenous effects altogether.
The results partially fall in line with the reviewed research in that past connectivity does affect new ties and mostly affects them positively. However, some findings either contradict or could not be hypothesized from the current state of knowledge. In particular, while most reviewed papers study how past connectivity of one type leads to new ties of a different type between the same two firms, the analysis finds that a new tie between two firms is significantly affected by their past ties to others. The presence and magnitude of effects depend on the type of the predicted tie, with alliances being least and executive migration—most affected by existing network structures.