Supplier Diversification Under the Threat of High-Impact, Low-Probability Disruptions

This paper investigates decision making in the context of sourcing in the presence of high impact, low probability supply disruptions. We develop a simple theoretical model that captures one of the central tradeoffs for sourcing decisions – while a consolidated (diversified) supply base reduces (increases) transaction costs, it increases (reduces) the exposure to costly supply disruptions. We predict, and using a controlled laboratory experiment, find evidence for, an oscillating sourcing pattern – decision makers temporarily increase their supply base immediately after a severe disruption, followed by reductions during stretches without disruptions. As we investigate decision making in a stationary environment, this pattern constitutes a costly bias. Furthermore, we find that decision makers tend to under-diversify their supply base on average. This result is suggestive of the idea that, in the long run, supply disruptions remain latent in the background of the decision problem. As low-probability disruptions almost never happen, the perceived value of a diversified supply base is rarely reinforced.