Network centrality, learning capacity and firm performance in equity alliance portfolios

Abstract

PurposePrior research from transaction costs economics argued that central firms perform better because they have superior access to information to discipline their alliance partners. Central firms may also, however, face higher costs and risks of unintentional learning and weaken their competence through structural inertia. We propose that these costs and risks are influenced by the learning capacities of the firms in the network and can explain different outcomes for focal firm performance.Design/methodology/approachTo test our predictions, we use instrumental variable–generalized method of moments estimation techniques on 15,517 firm-year observations from equity alliance portfolios in the global food industry across a 21-year window.FindingsWe find support for our predictions and show that the relationship between network degree centrality and firm performance is negatively influenced by partners’ learning capacity and positively influenced by focal firms’ learning capacity, while firms with low network degree centrality benefit less from their learning capacity.Research limitations/implicationsFuture developments in transaction cost economics may consider partner and focal firms’ learning capacity as moderators of the network degree centrality – firm performance relationship.Practical implicationsIn alliance decisions, managers must consider that the combination of high network degree centrality and partners’ learning capacity can lead to high costs, risks of unintentional learning, and structural inertia, all of which have negative consequences for performance. In concentrated industries where network positions are controlled by a few large firms, policymakers must acknowledge that firms may face substantial barriers to collaboration with learning-intensive firms.Originality/valueThis study is the first to develop and test a comprehensive transaction cost analysis of the central firm’s unintended knowledge flows and structural inertia in alliance networks. It is also the first to incorporate theoretically and empirically the hazards of complex and unintended information flows on the relationship of network degree centrality to performance in equity alliance portfolios.