Organizations invest heavily in the “war for talent”, competing to secure the services of the best individuals in a given industry. By doing so, they overlook the fact that, in addition to human capital, their performance is strongly affected by social capital. In this paper, we draw from the social networks and organizational routines literatures to argue that social capital, in the form of team stability and relational legacy, has a positive impact on team performance. Results from a longitudinal study on basketball teams provide strong support for our hypotheses. In particular, our finding that team performance is affected more by social capital than by human capital has important implications for human resource practices and may lead to reconsider the basic premises behind the “war for talent”.

Social capital has been characterized in the literature as either closure or brokerage. However, there is little evidence on which form individuals in a network tend to accumulate social capital, partly due to the limited availability of longitudinal studies that simultaneously study the emergence of both brokerage and closure. In this work, we shed light on how networks evolve by examining individuals’ propensity to accumulate these two different types of social capital over time. Analysis of the evolution of a network over two years suggest that network members have a general tendency toward closure and aversion to brokerage, which emerges only when controlling for the dynamic interplay of dyadic and triadic local structures. These results provide insights into network evolution and the behavioral tendencies of network members, and warn against the possibility of spuriousness when mis-specifying network evolution models.

This paper has two distinct goals. First, it introduces the concept of multiplex embeddedness and discusses its implications for organizational performance. While network studies have traditionally focused on the impact that one type of relationship (such as communication, trust or knowledge transfer) has on relevant organizational outcomes, organizational networks are inherently multivariate: within the same network, each actor is simultaneously involved in many types of relationships, which all at the same time affect its choice of action. To capture this phenomenon, I introduce the concept of multiplex embeddedness, which is the degree of involvement of an organization in a multiplex network and investigate its effects on organizational performance under two different conditions, i.e. complete and partial multiplexity. While high levels of multiplex embeddedness in the complete condition are positively related to organizational performance, strong multiplex embeddedness in the partial condition have the opposite (i.e., negative) effect on performance. Second, I explore the role of space in moderating the relationship between multiplex embeddedness and organizational performance. Until now, few works have focused on the influence of space on the network-organizational outcomes link. I argue that since networks exist in space, it is critical to investigate its role in moderating the relationship between network embeddedness and organizational performance. Specifically, I focus on the influence of both physical space (where the locale is determined by the firm’s geographic location within its industry, for example within an industrial district) and institutionally created space (where the locale is socially constructed through an institution, such as in common membership in an industry association). Contrary to copious anecdotal evidence, I argue that being located in very ’thick’ contexts such as industrial districts and industry associations is detrimental to the positive effects on organizational performance associated with strong embeddedness in an interorganizational network. Results from a 12-month field study of a whole industry strongly support these claims and represent a call for additional work in the area of multivariate networks and for a revision of consolidated beliefs regarding the influence of locales on organizational performance.

Processes of classification play a key role in shaping individual action because expectations and evaluations expressed by audiences are often contingent upon the assignment of individual organizations to institutionalized categories. Because audiences are typically reduced to a single representative agent expressing an average evaluation, comparatively less attention has been dedicated to how individual acts of classification may be related and give rise to multiplex roles within organizational populations. We address this issue by exploring the network structure of organizational fields emerging from interconnected decisions of individual field members to assign each other to identity classes. Using data collected on an entire industry (75 ceramic tiles machine manufacturers located in Central and Northern Italy), we study how the interaction between organizational attributes and relational mechanisms affects a producer’s decision to assign different categorical identity labels to other members of the community. We find that the effect of organization-level variables on individual classification decisions vary significantly by type of identity label, while the impact of relations is less differentiated. We also find that similar processes of hierarchical ordering characterized by the absence of reciprocity and cycles operate across settings defined by different identity labels. Finally, we document specific ways in which identity labels concatenate to induce multiplex relational patterns across settings and show how including this multiplexity in our models has limited or no impact on the individual label assignment decisions.

The success of firms located in geographic clusters has been repeatedly linked to the embedded nature of their ties. While much has been written about the consequences of embedded ties, i.e. ties featuring an economic and a social component, relatively little is known about their antecedents. Our goal is to clarify what drives the emergence of embedded ties in clusters. Building on the work of Uzzi (1996, 1997), we clarify and extend the theoretical framework on the antecedents of embedded ties, proposing different mechanisms of embedded ties creation at the dyadic and triadic level. Specifically, we expect that both social and economic drivers contribute to the emergence of embedded ties, that social ties have a stronger impact than economic ones in this process, and that the effect of third party referrals on embedded ties is contingent upon the type of network under examination. Building on recent methodological advances in the analysis of networks evolution, we use longitudinal network data on a geographic cluster located in Northern Italy to test our hypotheses about embedded ties creation and change. Results support our hypotheses, providing additional insight on the relative importance of economic and social ties in the emergence of embedded ties and shedding light on the type of local structures determining the evolution of networks.

Among the understudied elements contributing to the success of communities of practice (CoPs), communication networks play a central role. Communication flows among CoPs participants allow them to share both codified knowledge and, through periodic contact with other members, to get glimpses of a stickier, more difficult to disembody type of knowledge, which inevitably rubs off due to repeated interactions. Hence, we argue that certain types of communication network structures are likely to favor members’ participation more than others. Specifically, the success and sustainability of CoPs over time are likely to be associated to the presence of individuals taking on specific roles in the community. Similarly, other types of structures can hinder the creation and diffusion of knowledge, therefore negatively affecting CoPs survival. In this paper, we start investigating the link between communication network structure and CoPs success and failure by analyzing how online CoPs (i.e., email lists) evolve over time. These groups represent vivid examples of locales where knowledge circulates among people with different degree of expertise on a given topic, who contribute freely their knowledge to the community. Our goal is to provide initial support for a more articulated theory of how networks influence knowledge creation and diffusion via CoPs, which will help both scholars seeking to better understand CoPs and practitioners interested in implementing and/or supporting CoPs in the contexts in which they operate.

While closure and brokerage have traditionally been considered opposite forms of social capital, recent contributions have started to shift this view toward a complementary framework. Taking this new perspective, our work focuses on how these two forms of social capital evolve over time. Specifically, we investigate how endogenous (i.e., brokerage, closure and exchange) and dispositional factors affect the emergence of social capital. Results show that brokerage thwarts closure, closure enables brokerage, and network entrepreneurs adopt a “seesaw” strategy for the accumulation of both forms of social capital. These findings provide additional insight on the evolution of social capital by showing how brokerage and closure affect each other and how network entrepreneurs seem to act strategically in accumulating social capital.

Using a network perspective, I investigate how organizations are influenced by the structure of relationships in which they operate, focusing my attention on the network of interorganizational trust. Trust has been found to be a critical element of interorganizational arrangements and here I investigate how the embeddedness in the network defined for such relationship affects organizations involved in it. In doing so, I try to distinguish the effect of embeddedness in the trust network on network performance at two different levels of analysis: among the actors that make up the network core (defined as those network members interacting more frequently) and between the network core and the network periphery. The goal is to understand whether network performance is impacted only by strong level of trust among the critical actors of the network, or if instead it is also important to have elevated trust levels between such central players and the rest of the network. That is, whether trust’s impact on performance is a generalized effect throughout the network or is instead localized, i.e at play only in certain critical parts of it. Results strongly support the hypotheses that the strength and the pattern of interorganizational trust relationships play a critical role throughout the network in creating different perceptions of performance for individual organizations participating in it.

Despite the widespread popularity of structuration theory as an explanatory framework for a variety of organizational phenomena, its use in empirical studies has been limited--partially due to the inherent difficulty in operationalizing the central tenets of structuration. This paper addresses this situation in three ways. First, it argues that the conceptual sophistication of structuration may be fruitfully integrated with the methodological sophistication of social network analysis. Second, it suggests this integration facilitates the incorporation of existing findings about agents’ communication behavior in explicating the link between micro-behaviors and macro structure, which leads to the emergence of a network. Third, it leverages recent methodological advancements in social networks analysis (i.e., actor-oriented models) that allow analysis of the emergence of a social network grounded in structuration. We examine our structuration hypotheses using longitudinal data on the communication network of the employees of a public works department, which was collected every other month over a two-year period. Results indicate that while factors external to a social network play a role in understanding the emergence of structure, the structures which emerge from the aggregate of individual actions in previous time periods do in fact predict current structures, and that the influence of these reifications of the structure is greater than that of the external factors. We suggest that this provides a test of structuration theory. The paper concludes discussing strengths and weaknesses of this study, and implications for future research on the relationship between structuration theory and social network research.

In a context of dynamically changing networks, it can be interesting to find out whether the network evolution process locks in on some sort of equilibrium over time. Generally, dynamic processes such as the evolution of a social network in a given group of actors can exhibit different types of stochastic equilibrium behaviour. In this paper, we distinguish three aspects: stationarity, cyclicity, and temporal homogeneity of the process. They represent, in this order, increasingly liberal equilibrium concepts. A stationary network process is one in which observed networks can be viewed as subsequent realizations of the same underlying stochastic network model (or ‘regime’). Cyclicity generalises stationarity by allowing for time lags before recurrence to a previous regime of the process. Temporal homogeneity, finally, refers to processes that are themselves not stationary or cyclical, but in which changes in the regime follow a time-invariant pattern. For the case of complete network data collected in a panel design, we propose procedures that allow detecting these equilibrium concepts, which are based on exponential random graph models as well as actor-driven models for network evolution. To illustrate the proposed methods, we apply the new procedures to a data set consisting of a series of bi-monthly measurements of an intra-organisational communication network, which is measured over a two year period in bi-monthly intervals. We can show that in this data set, overall communication (defined by presence versus absence of any nonzero communication intensity) is not in equilibrium according to any of the three concepts, while strong communication (defined by a median split of communication intensities) is.

Organizational scholars have shown that both weak and strong network ties have a considerable influence on a variety of organizational outcomes. Studies examining their antecedents are very hard to come by. In particular, there is a paucity of works examining the processes which lead to the joint emergence of weak and strong network ties and which consider structural properties of the network itself among the possible determinants of tie strength. In this work, we analyze a network over time in an attempt to determine what are the demographic, formal and structural antecedents of weak and strong ties.

The view we articulate in this paper is based on the claim that the impact of network structure on knowledge transfer is not limited to dyadic interaction in a single network. Understanding knowledge flows - and hence absorptive capacity - requires assumptions about patterns of local dependence connecting subsets of organizations across multiple networks. We address this issue by exploring how multiplex network structures affect knowledge transfer among members of organizational communities. In addition to directly considering the impact of different types of local structures on knowledge transfer, we extend this literature in at least three directions. First, by looking at multiplex structures (i.e., dyads and triads emerging from the concatenation of different types of relations), we go beyond studies that have demonstrated third-party effects in social settings defined in terms of a single type of relation. Second, we explore how identities conferred by expectations of other industry participants - rather than internal organizational attributes - affect knowledge transfer (Hannan, Polos and Carroll 2007). Finally, by conducting separate analyses of the structural properties leading to the transfer of knowledge to other parties (knowledge transfer proper), and those having a role in the acquisition of knowledge from other parties (absorptive capacity), we explicitly account for non-reciprocation (relational asymmetry) - a fundamental micro-mechanism underlying the observed hierarchical structure of inter-organizational fields. We address these issues using data that we have collected on multiple exchange and influence relations among producers involved in the manufacturing of machinery for ceramics manufacturing located in Italy. We specify empirical Exponential Random Graphs Models (ERGM) to test how a multiplex interorganizational network affects knowledge flows among these organizations (Snijders et al., 2006). Our findings illuminate fundamental aspects of the link between network structure and knowledge flows, and point toward new directions for research on the relation between organizational identities, network structures and knowledge flows in organizational populations, fields and communities.

Communities of practice (CoPs) have become very popular as effective means to create and transfer organizational knowledge, thus drawing the attention of scholars belonging to various disciplines, such as anthropology, strategic management and organization theory. Managers and consultants have also quickly capitalized on this phenomenon as an extremely powerful tool to leverage organizational capabilities. However, practitioners have often acted too quickly, attempting to build on initial understandings of the processes of knowledge acquisition and replication that lie underneath CoPs without a solid grasp of the underlying mechanisms that make them successful in different contexts. As a result, many of the attempts to implement CoPs based on quick replications and generalizations of single cases ended up in expectable failures. The goal of this study is to start investigating some of these underlying mechanisms by looking at the antecedents and consequences of CoPs across different organizational contexts. More specifically, we present a multiple case study where we compare different organizations to identify which ones of their characteristics and employees’ attitudes toward knowledge creation and sharing are associated with different CoPs structures. In addition, we also start to speculate on the link between different CoPs structures and their outcomes.