Failure Risk and Quality Cost Management in Single versus Multiple Sourcing Decision†
I thank Stanley Baiman, Eshien Chong, Farok Contractor, Raffi Indjejikian, Francine Lafontaine, Martin Lariviere, Pinghan Liang, Anne Neumann, Eric Noreen, Veronica Santarosa, Richard Saouma, Tim Schmidt-Eisenlohr, Larry Snyder, Steve Tadelis, and Veikko Thiele for their valuable comments on earlier drafts of the article. I also thank participants of the 2011 Econometric Society European Meeting (ESEM) at Oslo, Norway, 2011 Conference of the European Association for Research in Industrial Economics (EARIE) at Stockholm, Sweden, 2011 World Congress of the International Economic Association (IEA) at Beijing, China, 2011 Accounting Research Workshop (ARW) at the University of Fribourg, Switzerland, 2011 International Industrial Organization Conference (IIOC) at Boston, Massachusetts, 2010 European School on New Institutional Economics (ESNIE) at Cargese Institute of Scientific Studies in France, and 2009 (Third) European Risk Conference on “Risk and Accounting” at Deloitte's training center in London for their useful feedback. All remaining errors are mine.
ABSTRACT
The advantage of multiple sourcing to protect against supplier failures arising from undependable products due to latent defects is examined using a model with nonlinear external failure costs. Prior research has focused only on supplier failures arising from unreliable supply, such as late, insufficient, or no delivery. I derive a closed-form characterization of the optimal production quota allocation for the LUX (Latent defect-Undependable product-eXternal failure) setting. The allocation determines the optimal supply base, with intuitive properties that hold under a mild requirement. The requirement includes the special case of equal procurement costs charged by suppliers but also allows unequal costs without any particular order. The key result of the article is a necessary and sufficient condition determining whether single or multiple sourcing is optimal. Another condition is obtained to determine the exact size of the optimal supply base, provided the mild requirement holds. With minor modifications, the results also hold when a buyer-initiated procurement contract can be used to elicit private information on the suppliers’ unit variable production costs.
