Facility Network Design with Equilibrium Outcomes

This article will contribute to the literature on facility network design by considering the role of competition in the evaluation of capacity investment decisions. As firms have turned to global supply networks to satisfy multiple regional demands, strategic decisions concerning the location and centralization of production must be made to balance a number of important factors: differentials in the cost of labor and inputs, proximity to key markets, and uncertainty in economic and regulatory conditions all play a critical role. Furthermore, responsive pricing and transhipment of products have enhanced the economic linkages between regional markets, making them increasingly difficult to analyze in isolation. We are motivated by problems of network design in capacity-intensive industries, such as chemical processing, technology and auto industries. In such industries, capacity investment represents a large portion of costs, and the economics dictate that resources, once brought online, must be used at a high rate. Such industries are especially vulnerable to market uncertainties, and overinvestment in capacity can spur intense competition, with the subsequent drop in prices making up-front investments difficult to recoup. Such price depressions have famously been observed in the paper, hard-drive, and auto industries, among others. Globalization offers one measure of protection against such uncertainties by facilitating the pooling of production capacity for use in serving multiple markets, and thus reducing the associated risks. Dong et al. (2010) cite the example of the chemical manufacturer Solutia, who, after making investments for its U.S. and European markets, uses transhipment between the two markets to balance out its mismatched demand. If the cost of transhipment is not too dear, then this flexibility will reduce risks and further enable the centralization of production to where it can be done most profitably. While there has been much focus on flexibility in the design of supply networks (see Kouvelis and Su (2005) for an overview of various design frameworks), it remains to incorporate the effect of such flexible production strategies on global market conditions. When one considers that optimal plant locations and capacity investment decisions are linked directly to the relative profitability of regional markets, such analysis becomes crucial. There are two important dependencies here to address: