برون سپاری داخلی در مقابل خارجی : شواهدی در سطح بنگاه دال بر نقش فن آوری

The decision about where to outsource varies across firms and industries. General machinery heavily depends on domestic subcontractors, while outsourcing overseas is prevalent in apparel. Based on firm-level data explicitly distinguishing foreign outsourcing from domestic outsourcing in all manufacturing industries, this paper finds that firms tend to prefer domestic outsourcing to foreign outsourcing when they are R&D-intensive. This finding is consistent with incomplete contracting models, since technologically complex products are likely to require high-quality contracting environment and assembler–supplier proximity. This paper also finds that firms connected with computer networks are actively outsourcing.

Outsourcing is increasing its importance in a wide range of firms and industries. While many assemblers have long been outsourcing to subcontractors in proximate locations, outsourcing across borders has been facilitated by information technology development and trade liberalization in recent years. Outsourcing to low-wage countries appears particularly prevalent in firms producing apparel and other labor-intensive goods, as in Nike, the company outsourcing most of its manufacturing tasks across borders. On the other hand, assemblers of technologically complex products tend to depend heavily on domestic suppliers. For example, Toyota employs the just-in-time procurement of customized auto parts from nearby suppliers. The relationship between technology and foreign outsourcing is worth investigating. The firm's decision about where to outsource has recently been formalized by the incomplete contract theory. For example, in the model by Grossman and Helpman (2005), the production is supposed to require a customized input manufactured by an independent supplier with a relation-specific investment, governed by an incomplete contract. A final assembler searches for a partner supplier close to its input requirement, either in the technologically and legally advanced home country North or in the low-wage foreign country South. Their model predicts that a firm tends to choose domestic outsourcing (DO) rather than foreign outsourcing (FO) when the firm's product is more technologically complex. The factors behind this choice can be listed as follows. First, customizing inputs is likely to be especially costly for Southern suppliers, many of whom are not equipped with computer-aided design (CAD). Second, these complex products tend to require high-quality legal system for verification/contracting. Third, the assembler-supplier proximity is often critical in customizing complex products. Finally, input markets for these technologically advanced products tend to be thicker in North. Many other recent theoretical models of incomplete contracting are also in line with this prediction (e.g. Acemoglu et al., 2006, Antràs, 2005, Antràs and Helpman, 2006 and Feenstra and Spencer, 2005).

This paper has investigated the firm's choice between foreign and domestic outsourcing, based on a firm-level data set covering 118,300 firms in all manufacturing industries without any firm-size threshold. The empirical findings are consistent with the theoretical prediction. R&D intensity is found negatively related with the extent of foreign outsourcing relative to domestic outsourcing. This result is one of the earliest direct evidence on the relationship between technological complexity and the choice of outsourcing location at the firm level. To cement the generality of this finding, however, several attempts will be desirable in future independent studies. At the aggregate sector level, intermediate import data derived from input–output tables can be combined with R&D data. One will also find it interesting to seek micro data on foreign outsourcing of non-production services, linked with data on human skills at the firm level, preferably in a longitudinal form.