New school speech regulation as a regulatory strategy against hate speech on social media: The case of Germany's NetzDG

Abstract

Germany's 2017 NetzDG law is an example of ‘new school speech regulation’ (Balkin, 2014), which restricts speech by coercing intermediaries into censoring users, rather than coercing speakers directly. It is the first such measure which specifically targets hate speech on social media, by requiring large platforms to operate complaints procedures which ensure illegal content is rapidly removed. Numerous other countries have since adopted similar regulations, indicating that states increasingly turn to new school speech regulation as a regulatory strategy to tackle hate speech on social media. This paper aims to evaluate the effectiveness of new school speech regulation in as a regulatory strategy to address online hate speech, taking NetzDG as a case study.A review of relevant empirical literature shows that many features of social media platforms actively promote or encourage hate speech. Key factors include algorithmic recommendations, which frequently promote hateful ideologies; social affordances which let users encourage or disseminate hate speech by others; anonymous, impersonal environments; and the absence of media ‘gatekeepers’. In mandating faster content deletion, NetzDG only addresses the last of these, ignoring other relevant factors. Moreover, reliance on individual user complaints to trigger platforms' obligations means hate speech will often escape deletion. Interviews with relevant civil society organisations confirm these flaws of the NetzDG model. From their perspectives, NetzDG has had little impact on the prevalence or visibility of online hate speech, and its reporting mechanisms fail to help affected communities.NetzDG represents an incremental, narrow approach to a complex sociotechnical problem which requires more fundamental regulatory reform. In this regard, it shows the limitations of censorship-based new school speech regulation. Rules which assert state authority by prescribing censorship of narrowly-defined content categories are ill-suited to large-scale, networked, algorithmically-curated social media, where other governance mechanisms influence user behaviour more than content deletion. The paper advocates a more systemic and preventive regulatory approach. Platforms should be required to take public interest considerations into account in all design and governance processes, aiming to shape platform environments to actively discourage users from posting or viewing hate speech, rather than simply deleting it afterwards.