This study aims to evaluate the impact of Japan’s 2022 online hate speech law on toxic content on digital platforms. Specifically, this work investigates whether the legislation led to a measurable change in the prevalence and intensity of toxic online content on Twitter and 5ch. This research seeks to contribute to broader discussions on digital governance, policy deterrence and content moderation in the context of rising global concern over online hate speech.
A natural experiment methodology was used, leveraging a large-scale dataset of over 16 million tweets and 100,000 posts from the anonymous forum 5ch. The analysis used Google’s Perspective API to quantify the toxicity of user-generated content before and after the enactment of the law. Comparative analysis was conducted across platforms and user types, focusing on changes in toxic volume and intensity.
Contrary to the legislation’s goals, there was no statistically significant reduction in toxic content. On Twitter, both the frequency and severity of toxicity increased post-law, especially among repeat offenders. Meanwhile, 5ch displayed no notable change in toxic behavior. These findings suggest that the law does not deter the production of toxic content and may have had unintended consequences on certain user behaviors.
This study offers one of the first empirical assessments of Japan’s 2022 hate speech law using behavioral data from major platforms. It challenges classical deterrence theory and provides novel insights into the limitations of blanket legal measures. This research underscores the need for context-aware, adaptive policy frameworks that reflect platform-specific dynamics and user behaviors in digital environments.
