Platform liability deters bad actors and increases proactive enforcement – platforms have the tech to implement it now

The model, while very simple, underscores several key insights. First, if firms have sufficiently deep pockets, and are held fully accountable for the harms they cause, then platform liability is unwarranted. Holding the firms (and only the firms) liable deters the harmful firms from joining the platform and interacting with users. If firms are judgment proof and immune from liability, however, then platform liability is socially desirable. With platform liability, the platform has an incentive to (1) raise the interaction price to deter the harmful firms and (2) invest resources to detect and remove the harmful firms from the platform. Moreover, platform liability can stimulate user participation when users are heterogeneous. To prevent overinvestment in detection and removal, the residual liability assigned to the platform may be partial instead of full. The optimal level of platform liability also depends on the intensity of platform competition, and whether users are involuntary bystanders or voluntary consumers of the firms. With appropriate incentives, platforms can play an important role in reducing social costs.110

Although internet platforms provided the motivation for this paper, our insights apply more broadly. Our analysis provides a strong economic rationale for holding traditional newspapers liable for harmful advertising content111 and for holding bricks-and-mortar retailers liable for the harm caused by defective products.112 However, we believe that the insights are particularly salient for online platforms including Facebook, Google, and Amazon. First, the harmful participants on platforms are frequently small and judgment proof with insufficient incentives to curtail their harmful activities. Second, the big tech giants have the data and technologies to detect and block participants that are more likely to harm others.

Our basic argument for holding platforms liable is valid regardless of the accuracy of the platforms’ current screening technologies and existing moderation efforts. First, the lack of effort by some platforms could reflect the weak incentives provided by the legal, economic, and political systems.113 Platforms may even have “perverse incentives” to

reduce their control of online activities, similar to the potential distortion caused by vicarious liability on organizations.114 Second, our model shows that platform liability may be socially desirable even if auditing is very costly or completely ineffective at detecting bad actors. Although platforms would not engage in auditing in this case, liability would force platforms to internalize the social harms and create an incentive for them to use the price mechanism to deter bad actors.