A number of management researchers have turned to examining organizations as networks of dynamic relationships. This view offers a mechanism to enable managers to analyze the processes of interaction unfolding in their organizations and to think about the real complexity of organizational life. The dynamic perspective may also provide easier solutions to problems. It is sometimes easier for managers to change one or a few network patterns in the interest of increased efficiency or improved attitudes than to shift entire departments around, as one would have to do in manipulating organizational charts.
Analysis of the static structural elements tends to focus on the organization as a whole, to take a long-range view. A good deal more is added in network research. One can look more closely at individuals or subunits making up various networks and translate organizational issues of centralization into individual issues of centrality. One can specify in terms of interactions the various roles people play in networks. And one can move relatively easily from thinking about organizations as totalities to thinking about various individuals in those organizations.
A network is a set of linkages among a defined set of people in which the character of the linkages is specified. Thus, a network may be built around job requirements or how best to get things done. It may be structured by social interactions or how people interact informally. Network observations began in laboratory research studies in the 1950s but were not done in real organizations. Groups of three, four, or five persons were studied to discover how variously imposed structures influenced problem solving and member attitudes. Structure was varied by imposing rules about who could talk to whom. The findings from these studies were consistent: for simple problems, centralized networks, in which information about the problem is sent to just one person, produced solutions faster with fewer errors than did decentralized networks, in which information was sent to everyone.
However, when problems were complex,decentralized networks were superior. Unfortunately, findings from laboratory investigations such as these do not always generalize to people at work.
For many years the results of these small-group network studies were presented in management textbooks as the gospel about how to engage in efficient problem solving. The research exhausted itself, however, and a number of years passed before interest was renewed in organizations as net works of relationships. One reason for the stagnated research was that larger organizational networks could not be studied. Yet these reflected the complex interrelationships found in reality. Complex networks were ignored until the 1970s because reducing data from them to understandable forms required high-speed computers that were not readily accessible. Today a number of computer programs are available for describing networks.
One kind of network of interest to managers is the grapevine. Grapevines are naturally occurring net works that are familiar to all members of organizations. Efficient and fast, grapevines are an avenue for managers both to obtain information about what’s going on in their organizations and to send out important information. Political leaders understand the value of pretesting their constituents’ acceptance of new programs and plans by leaking them to large-scale networks. Grapevines offer informal ways for managers to move information for any of a variety of purposes.
Various networks coexist in organizations and are used for purposes other than moving information. The three types, aside from the grapevine, are task networks, authority networks, and social networks. All the types overlap and serve different organizational functions. One set of high- technology military organizations over a six-month period found that task networks developed more quickly and became stable sooner than other kinds of networks. These were closely followed by the development of social networks. Authority networks were much slower to develop and never reached the level of maturity of the other two kinds. These findings are somewhat surprising, particularly for military organizations. They suggest that managers should pay close attention, in particular, to the development of task networks, implementing change where these networks appear to be dysfunctional to the goals of the organization.
A number of network properties have been identified. Although it is not difficult to infer some of the consequences of these properties for organizations, little research relating them to organizational performance has been done. These network properties follow:
– Connectedness, or the extent to which people in networks are interconnected
– Centrality, or the degree to which network relations follow the formal organizational hierarchy
– Reciprocity, or the degree to which there is two-way communication
– Vertical differentiation, or the degree to which different organizational levels are represented in the network
– Horizontal differentiation, or the degree to which different job areas are represented in the network