The dramatic growth in international outsourcing has fueled the expansion of international trade and deepened international integration. (1) Developments in outsourcing attract public attention due to concern that improvements in information technologies enable companies to relocate production or assembly activities to lower-cost overseas locations. (2) However, while international outsourcing allows firms to take advantage of factor price differences, Grossman and Helpman (2005) demonstrate that country cost differences are not sufficient by themselves to guarantee outsourcing, as international outsourcing decisions are based on many factors including the costs of search and customization.
More generally, new research in international trade recognizes information as a key element in the globalization process. However, informational improvements, such as reduced communication costs or increasingly sophisticated bar code transmission and management tools may not affect all producers uniformly. As Ranch and Trindade (2003) show, product differentiation in consumption or production, when combined with informational uncertainty, produces a degree of "natural protection." For this reason, information technology improvements are likely to deliver the greatest increases in global integration for those markets where the matching of differentiated partners is critical to the formation of new international partnerships. Similarly, as noted by Antras (2003) and Bernard et al. (2010), the organization of a firm's global operations depends further on factors including the importance of headquarters investments and the degree of product contractibility. (3)
A key goal of this paper is to study whether outsourcing decisions are consistent with outsourcing models that feature search and adaptation costs. To this end, this paper studies U.S. outsourcing conducted under the auspices of the overseas assembly program, or OAP. The OAP program was designed to allow firms to avoid tariffs on U.S.-origin inputs that they utilized in final goods they assembled overseas. Thus, while the OAP does not include all U.S. outsourcing activities--it does not capture U.S. assembly of foreign parts, or overseas contract manufacturing that is based on U.S. designs and specifications--OAP imports provide insight into a wide swath of U.S. outsourcing and represented 8.5% of U.S. import value during the sample period. Since the OAP provides tariff benefits to manufacturers who use U.S. inputs in their overseas production, it is administered by U.S. customs. Consequently, while the OAP provides insights into manufacturing outsourcing decisions, it does not provide information on developments in services outsourcing.
Since OAP exempts U.S. inputs from tariffs, the program requires participating firms to provide detailed information on the U.S. and foreign content of their outsourcing imports. Thus, OAP data records facilitate the precise measurement of cost shocks and trade frictions at the country-product level. In particular, heterogeneity in input choices both across countries and products can be exploited to identify cost pass-through and general price responses for outsourcing imports. For example, year-by-year pass-through can be estimated by regressing yearly OAP price changes at the product-country level on product-country cost changes. Plots of the year-by-year coefficients in Figure 1 show that the raw rate of OAP pass-through ranged from 6% to 23%. (4) However, accurate outsourcing pass-through estimation requires a theory-based estimation framework that controls for economic factors that also influence price decisions.
This paper adopts Feenstra's (1989) passthrough model, which is modified to account for the cost structure and tariff treatment facing outsourcing firms who produced products for import under the OAP. This framework, which is based on a Bertrand model of competition in internationally differentiated goods, focuses its attention on cost pass-through and on the degree to which producers emulate the price changes of their competitors. …