Disintermediation Governance and Complementor Innovation: An Empirical Look at Amazon.com

Abstract

This study investigates how the governance policy of disintermediation, i.e., restricting communication channels through which complementors can persuade buyers to circumvent the platform and transact directly, affects complementors' innovation behaviors. We leverage a governance policy change on Amazon.com (the focal platform) to examine complementors’ responses in adjusting innovation behaviors and product portfolios. Based on a coarsened exact matching (CEM) with the difference-in-differences (DID) empirical approach, we find that complementors who are affected by the disintermediation governance policy (treated complementors) significantly decrease the number of new products launched on the focal platform compared to the matched complementors who are not directly exposed to the policy (controlled complementors). Such an effect is mitigated by a complementor’s reputation and strengthened by a complementor’s capacity constraints. Supplemental analyses of the mechanisms show that treated complementors first significantly reduce the total number of new product developments, i.e., the income effect, and then increase the number of new products launched off the focal platform after the policy change, i.e., the switching effect. Moreover, treated complementors strategically switch the high-end products away while leaving low-end products on the focal platform, i.e., the bait effect. Our analysis of complementors' strategic responses to the platform governance policy against disintermediation sheds light on platform governance in balancing complementors’ extensive-margin and intensive-margin innovative activities and in directing complementors' innovation and product portfolio strategies.