Network curse effects: research based on listed manufacturing enterprises in China

Abstract

PurposeAlthough the importance of network embeddedness attracted much attention in recent years, the interpretations of the underlying mechanism almost focus on the positive effects and neglect the potential negative aspects. This paper aims to use Chinese listed manufacturing enterprises, in the perspective of network embeddedness, to analyze whether resource curse effects exist in strategic networks of enterprises.Design/methodology/approachThis paper examines the resource curse effect that exists in enterprise networks through analysis of enterprise’s total factors productivity compared with those of its industry peers with different degrees of embeddedness. This paper then uses several different methods, such as clustered fixed and random effect model, core and peripheral model, and generalized method of moments estimation in endogenous checks, to detect network curse effects and reveal three potential mechanisms, including overinvestment, extra maintaining cost and innovation extrusion.FindingsThis paper finds that excessive network embeddedness hinders continuous improvement of productivity, which is derived from three mechanisms: excessive network embeddedness makes information redundancy among enterprises, and the imitate-follow effect makes excessive investment seriously; enterprises have to encounter more problems and difficulties, such as information coordination and benefit negotiation, to maintain network relationship, which leads to extra cost; enterprises with abundant network resources may tend to more emphasis on marketable operational abilities, resulting in resource decentralization and less investment on innovation, which is finally not beneficial to productivity improvement in the long run.Originality/valueCompared with previous literatures, this paper not only enriches the understanding of negative mechanisms in enterprise`s network embeddedness, but also unfolds the inconsistency of network embeddedness effects in practice and finds new evidence that the two types of network embeddedness are consistent in certain circumstances.