Social Network Construction and Competitiveness in Global Value Chain Network Analysis Perspective on Chinese Aerospace Industry

Last categories of control variables take the firm’s embeddedness into consideration. The embeddedness represents the social and historical effect on economic life in terms of social structure, cognition, politics and culture (Uzzi, 1997). In terms of formation of firms’ business networks, social embeddedness is reflected in a path-dependent adaptation and reciprocal commitment between business partners (Andersson, Forsgren, & Holm, 2002). In this study following variables are included as control for social embeddedness:

There are few issued to clarify ahead of the empirical study. First, due to the data availability and sensibility, only non-military production and services of investigated units will be included. In addition, the database build-up is primarily based on secondary data available to the public and the main purpose of the research is not to exhaust all types of linkages in Chinese aerospace industrial networks, therefore, only formal business linkages that are clearly stated in the original sources are included. The author realize the important of informal linkages between units, especially in emerging economies. Nevertheless, to include informal linkages, it requires extensive qualitative investigation such as survey and interviews, which exceed the original purpose and resources allowed in this research. In this sense, the author suggest further research could conduct ethnographic observation and survey to understand how informal interpersonal relationships can affect the strategic position and performance of different types of units in industrial networks.

The dependent and independent variables introduced in this study encompass both nominal variables (network position, geographic location, and embeddedness) and continuous variables (eigenvector centrality). Multinomial logistic regression analysis solve this problem of compatibility of variables. For each category of nominal variable, one group of parameter will be selected as base group, while the regression coefficients of other variables within the same category will be calculated based on logit function. For category of value chain position, we select firms specialised in “related industry” that are neither engaged in primary activities nor support activities as base. For geographic locations, domestic firms that are not located in output-intense provinces is selected as based. Embeddedness control variables are listed in their original form since unlike the previous two categories, they are not exclusive to each other.

In supply chain network, firms specialised in primary activities stringently follows the tier-based hierarchy. Those allocated at the later stage of the value chain have better off network position in comparison to those allocated at the forward stage. H3 is supported in supply chain network. However, the regression of coefficient of firms specialised in support activities (0.24) lie between OEM and multiple-tier suppliers. It is predicted that although firms specialised in support activities do not have the same level of influence and control power in supply chain networks in comparison to OEMs and customers, nonetheless, thanks to their role as knowledge distributor and relational coordinator, they may still have better network position than suppliers in the supply chain network. Therefore, H2 is not fully supported in supply chain network.

Regarding geographic locations, we can observe the divergence between partnership network and supply chain network as well. Since neither of the two categories of foreign origin turn out to be significant at the same time in either partnership network or supply chain network, H4 lacks empirical support in both networks. Nonetheless, if we compare foreign firms with domestic ones, prominent findings can still be concluded. In partnership network, origin of developed economies does not have significant impact on firm’s network position, while firms from other emerging economies have better off position than domestic firms located in output-intense regions but have less advantageous position than domestic firms located in output-intense regions. In supply chain network, the impact of origin of other emerging economies is not significant. Nonetheless, firms from developed economies have better off position than both types of domestic firms. Regarding domestic firms, it turns out that the influence of location in high output efficiency regions overwhelms the rest of the country in both partnership and supply chain network, thus H5 is strongly supported in both types of networks.

Last but not least, from the regression result of embeddedness control variables, we may conclude that, although social embeddedness also influence firms’ network position, nonetheless, the contributions of different sorts of embeddedness attributes also diverge depending on the type of network. In partnership network, longer duration of market presence, coreness of industry in the network and identity as headquarter all significantly contribute to firms’ position in partnership network. While in supply chain network, only co-existence in both types of network significantly contributes to the network position, In addition, similarity in legal system does not significantly affect firms’ position in both types of networks.