Résumé : In the realm of globalization, international sourcing of services contributes to reshape firm’s value chains as the physical dispersion of these activities increases. This reorganization does not simply lead to the replication of domestic activities in a destination providing resource advantages, but, in most cases, it implies profound modifications of the flows of activities, including the reconsideration of the boundaries of the firm. Global sourcing strategies, also called offshoring, seek to increase firm’s efficiency by combining the exploitation of foreign locational advantages with process redesign. When aggregated, these firm-level strategies translate into considerable international exchanges to a point that flows of intermediate services represent about 73% of the total of international trade in services for 2005 (OECD, 2009). These activities present a high degree of heterogeneity in terms of functions concerned, the related domestic industries, motivations, destinations, organizational structure or scope. This wave of internationalization, because of its relative novelty, growth and rapid diversification, draws the interest from the public, political and academic spheres but the comprehension of the determinants shaping the configuration and organization of these activities still remain largely unknown. Based on four essays, this PhD thesis addresses the impact of host-country characteristics and distance factors on the configuration of international sourcing activities in the dimensions of location, governance model and scope of activities.

The first paper studies the country-specific determinants of the interdependent choices of destination and governance model in the global sourcing of services. I explore the simultaneity of these decisions and I jointly estimate their determinants using implementation-level data. Derived from comparative advantages, host-country uncertainty and the global dispersion of tasks, I present three classes of factors driving global sourcing configurations: resource arbitrages, host-country risk and communication barriers. Empirical results confirm that locations with resource or capabilities advantages specific to services – low labour cost, education and labour supply – attract more offshoring activities. However the pursued resource advantages differ depending on the governance model. Country attractiveness for captive implementations presents a higher positive sensitivity to the education-intensive resources, while outsourcing strategies have a greater cost-cutting orientation coming from labour cost arbitrages. Furthermore, the risks inherent to the host-country, in the form of weak formal institutions and inexperience in the destination, have the dual effect of deterring location attractiveness, while they foster the adoption of the outsourcing model compared to the captive one. Communication barriers coming from geographic distance, cultural and linguistic differences have the simultaneous effect of discouraging global sourcing in those locations while, to overcome these constraints, firms favor higher integration with the use of captive models.

This second paper further explores the mechanisms through which home-host country distances affect the choice of governance mode in service offshoring. Using a Transaction Cost Economics approach, I explore the comparative costs of the hierarchical and contractual models to show that different dimensions of distance (geographic, cultural and institutional), because they generate different types of uncertainties, impact offshore governance choices in different ways. Empirical results confirm that, on the one hand, firms are more likely to respond to internal uncertainties resulting from geographic and cultural distance by leveraging the internal controls and collaboration mechanisms of a captive offshore service center. On the other hand, they tend to respond to external uncertainties resulting from institutional distance by limiting their foreign commitment and leveraging the resources and local experience of third party service providers. Finally, I find that the temporal distance component (time zone difference) of geographical dispersion between onshore and offshore countries plays a dominant role over the spatial distance component.

The third section then concentrates on the impact of the institutional environment (regulative) on international sourcing activities. To exploit country-specific advantages, firms that source activities from abroad are forced to integrate the institutional environment into the choice not only of host-country, but also of governance model for their offshore activities. Considering inefficient institutions as drivers of transaction costs, this conceptual paper explores the impact of the host-country regulative environment in the interdependent decisions of country selection and governance model (captive or outsourcing) in firms’ global sourcing strategies. I consider two classes of assets: transferred assets for knowledge/information flows, and local assets sourced from the host location. I show that each class involves specific institutional risks for offshoring practices. In turn, because of the different institutional exposures of the captive model and the outsourced one, the institutional risks associated with transferred and local assets have different implications for the choice of governance model. Firms react to institutional risks relative to transferred assets by internalizing their activity, but they bypass inefficient institutions for local assets using outsourcing. Based on the interaction of the institutional risks relative to each class of assets, I then obtain sufficient conditions that give the firm-optimal combinations of country selection and governance model.

The last section studies how firm-level and country-level risks affect the scope of the process operated in the foreign unit. To prevent appropriation hazard for proprietary content, firms choose a particular disaggregation of the value chain. We argue that, in response to the lack of control offered by internalization and the lack of protection provided by host-country institutions for protecting proprietary content, firms reduce the scope of their activities. In other words, they exploit existing complementarities between the tasks of their value chain using a higher disaggregation of their process and therefore reducing appropriation value for outsiders. Based on a sample of 750 international sourcing projects, regression results on the scope of offshore activities confirm that firms prefer to source discrete tasks rather than entire processes when they lack the protection of internalization and external institutions. In addition, experience modifies these relationships. On the one hand, inexperienced firms do not rely on this slicing mechanism to prevent the loss of control implied by an outsourcing model. On the other hand, the effect of weak institutional protection is perceived as more stringent for inexperienced firms. When host-country institutions are deficient, these firms, compared to the experienced ones, have a higher propensity to operate discrete tasks rather than entire processes.