The costs of increased localization for a multiple-product food supply chain: Dairy in the United States

Abstract

There is increased interest in greater localization of food supply chains but little evidence about the effects of localization on supply-chain costs. Assessing these effects is complex in multiple-product, multi-process supply chains such as the dairy industry. In this study, we develop a spatially-disaggregated transshipment model for the US dairy sector that minimizes total supply-chain costs, including assembly, processing, interplant transportation and final product distribution. We employ the cost-minimizing solution as benchmark to compare alternative scenarios of increased supply chain localization. Our results indicate: (1) short-run limits to increased localization, (2) modest impacts on overall supply-chain costs, and (3) large cost re-allocations across supply chain segments, regions and products. We find that increased localization reduces assembly costs while increase processing and distribution costs. Cost increases are larger in regions with smaller raw milk supplies and during the season when less raw milk is produced. Minimizing distances traveled by all dairy products results in tradeoffs across products in terms of cost and distance traveled. The relationship between increased localization and costs appears to be nonlinear.

Introduction

There is increased interest among consumers, private and public decision makers regarding the sustainability of food supply chains. Today, consumers are more responsive to the ways food is produced, processed and distributed, often based on perceived benefits, including environmental, health, food safety, and rural development. Policy makers, for their part, are pressuring food industries to re-examine the sustainability of their supply chains. The US dairy industry, for example, has a plan to reduce greenhouse gas (GHG) emissions by 25% by 2025 (IDFA, 2008). The initiative incorporates GHG emission reductions in production (crop and dairy farming) as well as in the supply chain (transportation, processing, distribution and retailing), which account for 70% and 30% of sector GHG emissions, respectively (EPA, 2008).

One consequence of the pressure to improve sustainability performance is the emergence of arguments in favor of more localized food supply chains. Advocates of increased localization argue that reduced GHG emissions are one possible benefit—among many potential ones—of localized food supply chains. As a result, concepts such as “food-miles” and the like often have been employed to develop metrics for evaluation of sustainability performance, primarily because they are relatively easy to measure and communicate to the public (Coley et al., 2009). Therefore, increased localization of food supply chains has become linked to (even synonymous with) the reduction of GHG emissions (Darby et al., 2008, Hein et al., 2006, Jones et al., 2007, Lang and Haesman, 2004, Peters et al., 2009, Stagl, 2002).

Although localization may be desirable, there is limited empirical evidence about how increased localization may influence costs of food supply chains and, ultimately, the price for food paid by consumers. The business model that has evolved in conventional, mainstream food supply chains delivers multiple benefits stemming from the provision of a wide variety of convenient, year-round, relatively inexpensive products (King et al., 2010a, King et al., 2010b). There is little knowledge about possible tradeoffs between increased localization and the cost of food supply chains.

Examining possible tradeoffs between increased localization and food supply costs in the context of multi-product industries requires spatially-disaggregated models that take into account the multiple relationships among the many supply chain segments beyond the farm gate, including assembly, processing, transportation and distribution. One approach that meets these analytical requirements is spatial optimization modeling. To analyze the impacts of greater localization on supply-chain costs, we employ a spatial optimization model of the US dairy product supply chain. The model focuses on supply chain segments beyond the farm gate (assembly, interplant transportation, processing and distribution) for all dairy products, of which the most important are fluid milk, yogurt, cheese, butter and nonfat dry milk. We calibrate the model using data from 2006 and develop scenarios to compare impacts of alternative strategies to increase localization: one focusing on reducing the distance travelled by all dairy products and the other focusing on the reduction of one product only (fluid milk).

The dairy sector is an excellent example for examining the economic consequences of increased localization of food supply chains. First, dairy was primarily a local industry in the US before 1950. Since then, rapid innovations in food preservation and processing, huge investments in private and public infrastructure, as well as the realization of important economic benefits accruing to economies of scale and specialization, have all contributed to a transition of the US dairy supply chain from local to national (and even global). Second, although milk is produced in every US state, there are significant spatial imbalances between production and consumption regions, and these differences have grown over time (Fig. 1, Fig. 2). The western US has experienced large increases in milk production but has a relatively low population density, while the southeastern US has gone through substantial population growth accompanied by shrinking milk production. Third, dairy is a multi-product industry with various interrelated supply chains and disentangling the consequences of increased localization efforts is not straightforward. Product diversity and the complexity of product flows in the dairy supply chain, together with the high level of perishability of raw milk and many of the intermediate and final products, make consideration of how to increase localization of this industry challenging.

This study is organized as follows. After this introduction, we discuss the literature on localization, emphasizing the links to supply-chain costs. Next, we describe our multi-product optimization model of the US dairy supply chain. In turn, we discuss our alternative scenarios, present our results and discuss the policy implications. The last section offers concluding remarks, discusses limitations of our study and proposes topics for future research.