Outsourcing, delocalisation and firm organization: transaction costs vs. industrial relationships in a local production system of Emilia Romagna

Abstract

Empirical evidence shows that both the volume and the value of intermediate inputs and business production services contracted out by firms, both at home and abroad, have risen dramatically in the last two decays. The determinants and the implications of outsourcing, in particular, that is of the “buy-rather-than-make” decision, have become a core topic in industrial organization. The attention of standard approaches has focused on transaction costs, ownership allocation and efficient investments, formal vs. real authority and, in general, on the entailed incentive conflicts. Outsourcing has also attracted the attention of ‘non-standard’ approaches, which have focused on production, rather than transactions, by addressing the role of firms’ capabilities and competences. In spite of the inner differences of these approaches, it has been recently argued that understanding vertical integration and disintegration could benefit from overcoming the “production-transaction dichotomy” their independent analyses imply. While sharing this point of view, in this paper we claim that a further effort of combined analysis is required in order to capture the “embedded” nature of the outsourcing firm. The paper aims at investigating how far transaction costs economics concurs in the explanation of outsourcing decisions in firms characterized by ‘thick’ industrial relations. The research hypotheses are tested with respect to a representative sample of manufacturing firms for a local production system in Emilia Romagna by matching two datasets built up by the authors through detailed surveys in 2002 and 2005. The empirical model specifies as dependent variables, at first, a synthetic index of outsourcing intensity deriving from the number of outsourced activities in 2004, regardless their functions within the firm. Then, specific analyses are carried out separately at the level of production activities, production support and ancillary activities. As far as the explanatory variables are concerned, they relate, first, to a kernel of TCE related drivers, secondly, to a bundle of diversified variables capturing various elements of industrial relations, and, finally, to a rich vector of control belonging to the area of structural firm features and techno-organisational characteristics. In addition, information on firm delocalisation strategies deriving from the 2004 survey allows an investigation of the extent to which delocalisation affect and interplay with outsourcing and its drivers. Delocalisation related variables are used for two purposes by using different empirical modelling: first, delocalisation strategies are added to the vector of controls in rode to verify the robustness of results at the light of such an inclusion, which may cause in this conceptual framework an omitted variable misspecification problem. Secondly, Two stages Heckit models are implemented to verify whether outsourcing dynamics differ between firms that delocalise production process and the rest.