FACULTY AND RESEARCH

Concerns about the environment, public health, and safety are driving governments around the world to restrict access to their markets of products, which may present environmental/health risks. As a result, legislation in many countries requires that firms comply with specific environmental and safety standards before introducing their products in these markets. However, there is no unified global standard, and a set of local rules governs most of the markets. Therefore, products complying with the regulations of one market may not be compliant with the standards laid out in the other. This means that large firms must include the local regulatory requirements of their targeted markets into account when making their sourcing decisions. The rise in the enforcement of compliance legislation has overlapped with an increase in the trend of global sourcing, where manufacturers have opted for a diversified and offshore supplier base. With the issue of compliance legislation in place, the suppliers who otherwise would produce functionally identical products become differentiable based on their ability to fulfill compliance requirements of the targeted markets. Also, the suppliers have incentives to exaggerate their compliance ability or may lack the technical capability to ascertain this information completely. Therefore, there always resides a degree of uncertainty about a supplier’s actual compliance capability. On the market side, there is uncertainty concerning product demand in individual markets. These two sources of uncertainty make this a rather challenging problem. We propose a formulation of supplier selection and product allocation decisions in this problem based on a two-stage robust optimization method. Through a set of numerical experiments, we discuss some exciting insights about the role of uncertainty in compliance failure on overall profitability and flexibility in the supply chain. We show that 3% supply-side uncertainty has almost the same effect on the quality of worst-case profit as 30% deviation from the nominal demand.