6 Policy Brief No. 45 Toda Peace Institute Real Name policy Facebook has long required real names on the platform, sometimes claiming that this policy increases civility. Indeed, for a while, it was fashionable for pundits to insist that all internet platforms ought to do this — yet there has long been opposing evidence. This 2012 TechCrunch article by Greg Ferenstein points out, “there’s surprisingly good evidence from South Korea that real name policies fail at cleaning up comments. In 2007, South Korea temporarily mandated that all websites with over 100,000 viewers require real names, but scrapped it after it was found to be ineffective at cleaning up abusive and malicious comments.” In other words, there is no solid evidence base that suggests that real name policies actually improve online discourse (but real name policies may serve other purposes, such as improving the company’s ability to build data-rich ad targeting profiles). This may be one of the least well-supported interventions any major platform has tried. Ad transparency In 2018, after coming under fire for hosting non-transparent advertising (which is especially worrying in a world drowning in propaganda and false news), Facebook made it possible to track all the ads a given Page is running when you look at the Page. This was hailed as a good first step, though many commentators felt it was not nearly enough. Ben Thompson at Stratechery, a comparatively tech-friendly blog, was critical when he wrote that “These ads can still only be seen by going to the actual pages, which are impossible to know about unless you are shown an ad; the company should have a central, searchable, repository of all those hundreds of millions of ads.” “Community Standards” transparency For years, it was difficult for people outside Facebook to learn about the company’s “Community Standards,” the internal guidelines that moderators follow to decide which content and accounts to remove. This non-transparency did not go over well with the public, so in 2018, Facebook made the Community Standards public (including the standards of what counts as sexual content, fake accounts, child exploitation, inappropriate violence, and other stuff that the company will censor). In early 2019, Facebook announced that it is building an outside “Oversight Board” to help decide its most challenging content moderation cases, but it is not yet clear what this Board will look like. Facebook also works with outside organisations to stay on top of fast-moving areas such as hate speech, where many people use coded language — i.e., the words that count as hate speech often change. These partnerships seem good in the sense that they add transparency and accountability to the process, but some critics say the platform is asking civil society partners to do work that Facebook should be taking on itself. Content moderation processes, at the implementation level Aside from the transparency (or lack thereof) of their guidelines, Facebook has also come under fire for outsourcing most of the content removal to consulting companies whose employees are far less privileged than Facebook’s core employees, as detailed in a 2019 Verge article by Casey Newton and 2018 documentary “The Cleaners” by Riesewieck and Block. These content moderators report both significant personal trauma (from looking