Platform Governance in the Presence of Within-Complementor Interdependencies: Evidence from the Rideshare Industry

Existing studies suggest that platform access restriction may cause restricted complementors to switch to competing platforms, which will increase complement quantity on competing platforms. We re-examine this prediction by accounting for the impact of economies of scope on complementor responses to platform access restriction. We argue that restricting a complementor's access on a platform may prevent it from achieving economies of scope from multi-homing, thereby incentivizing it to abandon both the restricted and (unrestricted) competing platforms. Using rideshare data in New York City, we compare the numbers of trips made by Lyft and Uber drivers, respectively, before and after Lyft restricted drivers' access on its platform. We find that Lyft's access restriction reduced trip numbers not only on Lyft but also on Uber. In addition, both Lyft's and Uber's trip numbers decreased not only during the restricted low-demand periods (e.g., non-rush hours) but also during the unrestricted high-demand periods (e.g., rush hours). In contrast, after a substantial number of multi-homing drivers left both platforms following Lyft's access restriction, a subsequent access restriction by Uber led to an increase in trip numbers on Lyft. These results highlight the importance of accounting for interdependencies across complementor activities when designing platform governance policies.