Social Responsibility in Supply Networks

Thesis (Ph.D.) - Indiana University, Kelley School of Business, 2020I study how the structure of supply networks interacts with efforts to make supply chains more socially responsible. One example concerns mineral mining in the Democratic Republic of the Congo that funds armed conflicts. Nonprofits and legislative bodies pressure manufacturers to trace and disclose their mineral sources. My first essay studies the decisions of manufacturers and smelters in the mineral supply network. We show the equilibrium depends on the total demand of “compliance- prone” manufacturers, who would comply if the prices of certified and noncertified metals were equal. Our results imply once penalties make sufficiently many manufacturers compliance-prone, certified metal may become so expensive that some compliance-prone manufacturers will not comply. Five companies established a common fund for auditing the mineral smelters. Since the list of smelters certified by the audits is public, companies have an incentive to free-ride. Despite this incentive the fund was a success and received subsequent contributions from dozens of other companies. My second essay studies why. We consider two factors: an early-stage alliance and status-seeking behavior. We model the funding initiative as a public goods game and test the results in laboratory experiments. Our experiments show that the invitation stage is key to high contribution and status-seeking behavior affects the forming of an alliance. My third essay studies a buyer auditing suppliers within a network to identify noncompliance. If a supplier fails an audit, the buyer must rectify the supplier or drop it (along with dependent suppliers). The network topology evolves as the buyer drops suppliers. We show the buyer should first audit and drop some suppliers, then either rectify all remaining ones, or proceed directly to production. When focusing on an upper tier, the buyer should always audit the least valuable unaudited supplier, yielding greater balance in the network structure. We establish the condition under which the buyer may truncate auditing (“hear no evil, see no evil”)