Network dyads in entrepreneurial settings: a study of the governance of exchange relationships - includes appendix

Administrative Science Quarterly, March, 1992 by Andrea Larson

Social control in network organizational forms is examined through an inductive field study of a sample of dyadic relationships established by high-growth entrepreneurial firms. The social dimensions of the transactions are central in explaining control and coordination in the exchange structures. A process model of network formation is presented that highlights the importance of reputation, trust, reciprocity, and mutual interdependence. The network form is proposed as an alternative to vertical integration for high-growth entrepreneurial firms. The data also suggest that studying the network form of governance can provide insights into firm growth.(*)

INTRODUCTION

The recent proliferation of network organizational forms that don't fit neatly into either the market or hierarchy frameworks proposed by Coase (1952) to explain economic exchange has resulted in some scrambling to explain how such organizations are governed. The traditional transaction cost approach also has not been very useful in understanding the governance of such organizations. According to the transaction cost approach (Williamson, 1975, 1985), the frequency, uncertainty, and specificity of assets dedicated to a particular transaction influence the costs associated with market contracts. When costs rise sufficiently high, markets fail and transactions are absorbed within the firm. Williamson (1991) has tried to come to terms with these network forms in recognizing what he calls the hybrid form of governance. Hybrids combine aspects of market transactions and characteristics of hierarchies and fall between the two alternatives on a continuum. While price incentives, legal contracts, and administrative monitoring play a role, cost minimization is the dominant control mechanism that explains the hybrid organizational form.

The growing number of empirical studies of a variety of hybrids, however, has challenged the cost-based view. These exchange arrangements alternatively have been called quasifirms (Eccles, 1981) and relational contracting (Macaulay, 1963), as well as hybrids (Powell, 1987) and networks (Powell, 1990). They have been described in a variety of settings, including international business (Contractor and Lorange, 1988), the U.S. construction and publishing industries (Eccles, 1981; Coser, Kadushin, and Powell, 1982), Japanese textiles (Dore, 1983), Swedish large manufacturing companies (Hakansson, 1987), and entrepreneurial firms (Johannisson, 1987; Jarillo, 1988; Larson, 1988; Lorenzoni and Ornati, 1988). In contrast to the efficiency and transaction cost explanation of control and governance, these accounts point to reciprocity norms, personal relationships, reputation, and trust as important factors explaining the duration and stability of the exchange structures.

As increasing numbers of these non-market and non-hierarchy forms of organization have been documented, Williamson's continuum notion of governance structures and the categories themselves have come under question. Powell (1990) has built an argument for the network form as a distinct organizational arrangement, as opposed to a hybrid form that combines market and hierarchy. He defined networks by a set of descriptive characteristics and critical components. They are distinct from market or hierarchical arrangements in their heavy reliance on reciprocity, collaboration, complementary interdependence, a reputation and relationship basis for communication, and an informal climate oriented toward mutual gain. in contrast, market governance arrangements rely primarily on price for control, and hierarchical structures rely heavily on administrative authority. According to Powell, firms more likely to engage in network arrangements will be those needing to exchange difficult-to-codify, knowledge-intensive skills that are best transferred through processes of collaborative information sharing. Powell also pointed to firms engaged in fast-moving industries with short product cycles as likely to engage in network partnerships in order to reposition products rapidly and respond quickly to changing market conditions and technological developments. Bradach and Eccles (1989) mapped price, authority, and trust as control mechanisms onto market, hierarchy, and network, respectively, and noted that these are ideal types that in reality are often combined. Such combinations have been documented by Stinchcombe and Heimer (1985), who found hierarchical dimensions of oil-exploration-project contracts, and Eccles (1985), who described market forces and hierarchical authority at work in transfer pricing within firms.

These alternative views on the nature of exchange governance point to the need for more information on how the network form is built and controlled. Existing studies offer pieces of the picture. In a variety of settings, investigators have reported social characteristics of network exchange structures. Personal relationships that resulted in allegiances were found to shape economic outcomes in the publishing industry (Coser, Kadushin, and Powell, 1982). Reciprocity, the importance of good personal relations, and trust-based cooperation provided control and coordination and enhanced the effectiveness of network forms in international joint ventures (Walker, 1988; Doz, 1988; Hakansson and Johanson, 1988). Looking at network structures formed by Italian small to mid-sized textile firms, Lorenzoni and Ornati (1988: 55) found "unconventional mechanisms of coordination" guiding the collaborative exchanges, including trust, reciprocity, and mutual adjustment. In Japanese networked firms, trust and mutual obligation enhanced information flows. These characteristics were reinforced by a concern for reputation to give stability and longevity to interfirm partnerships (Dore, 1987). In his study of strategic networks and entrepreneurial firms, Jarillo (1988) argued that trust and concern for reputation are important in sustained cooperative exchange structures. The findings suggest that these control mechanisms are important, yet they have never been explicitly studied.