Key points are not available for this paper at this time.
Purpose – The purpose of this paper is to investigate the performance implications associated with production outsourcing. Specifically, the paper analyzes the cost of goods sold for firms who outsource core manufacturing processes, using empirical data from a variety of industries. The paper seeks to better understand the influence of outsourcing on factory cost by looking at these in the context of related strategies, such as supplier integration, information technology (IT) implementation, and manufacturing process decisions. Design/methodology/approach – The paper draws on transaction cost economics, manufacturing strategy, and supply chain management literature to aid in predicting the performance to be expected when outsourcing production activities. Furthermore, the paper investigates the moderating effects of manufacturing strategies, supplier integration, and IT expenditures on outsourcing. The primary model is a two-way panel model for the cross-sectional and longitudinal data drawn from the MPI Census of Manufacturers Survey of US manufacturing plants. Findings – The analysis indicates that production outsourcing tends to shift costs among cost of goods sold (COGS) categories, but does not consistently reduce them as measured by overall COGS. The effects of production outsourcing on both the cost of labor and the cost of materials are strong, tending to decrease labor, and increase materials. Additionally, this study shows that a high level of supplier integration has a notable moderating influence on overall COGS, but that process strategies do not. Finally, this analysis indicates that IT expenditures were not influential as a moderator variable when outsourcing, but did have a marked influence on overall COGS, as well as on labor and materials costs. Originality/value – This research investigates the effects of outsourcing on the components of COGS, a level of analysis that is typically not looked at relative to outsourcing. This research also provides methodological contributions with the development of a nested random effects structural model for use with a secondary data source.
Meixell et al. (Tue,) studied this question.