All Failures are Not Made Equal: Towards a Comparative-Contingent Theory of Learning from Failures

Abstract

Although much scholarly work has investigated how organizations learn from past failures, we still lack knowledge about whether and how different types of failure experiences interact and jointly help firms prevent future failure events. This article develops a novel comparative-contingent theoretical framework that considers firms’ experience with both small and large failures, as well as experiential and vicarious learning mechanisms through firms’ both direct and indirect failure experience. We test our hypotheses on an original dataset comprising all companies that have issued high-severity and low-severity medical device recalls in the decade 2009-2018. We provide evidence of a strong experiential learning effect from both small and large failures, which however is salient only to prevent the same type of failure but not the other. We also find that firms’ learning from small failures is contingent on two experiential factors. Firms directly experiencing a high number of large failures prove unable to learn from their own small failures, while exposure to a high number of peers’ small failures reduces firms’ learning rate from small failures. This article contributes to the scholarship addressing organizational learning by demonstrating a differential salience of small vs large failure events in stimulating learning and a detrimental interaction of experiential and vicarious learning mechanisms.