来源：《华盛顿邮报》
刊登日期：2021年4月2日
The 2020 election brought with it a furious controversy over digital political advertising: whether social media sites should fact-check ads, to start, and later whether they should allow ads at all. A new Duke University paper reveals how uninformed this fight was, as well as what changes are necessary to ensure politicians, platforms and more know what they’re doing next time.
2020年的大选带来了一场关于网络竞选广告的激烈争议：首先，社交媒体网站是否应该对竞选广告进行事实检查，然后，是否应该允许竞选广告。杜克大学的新论文揭示了这场斗争是多么的无知，以及为了确保政客、平台和更多的人知道他们下次做什么，有必要做出哪些改变。
Former president Donald Trump’s tendency for bald-faced lies provoked demands that Facebook and its peers refuse to accept payment for promoting them. The ultimate outcome was bans on all political advertising by Twitter, Pinterest and several other sites, as well as by Facebook and Google after the general election. But whether these measures actually resolved the problem of misinformation remains uncertain. So is whether they introduced different problems of their own, from the suppression of speech to the inability to spread good information.
美国前总统唐纳德·特朗普喜欢赤裸裸地撒谎，这促使Facebook等平台拒绝接受为宣传这些谎言而支付的报酬。最终的结果是，在大选投票结束之后，Twitter、Pinterest和其他几个网站，以及Facebook和谷歌禁止了所有的竞选广告。但这些措施是否真正解决了虚假信息的问题，仍不确定。同样不确定的是，从压制言论到无法传播好的信息，Facebook等平台是否是这些新问题的始作俑者。
Yet, answers to those questions are hard to come by even today. The Federal Election Commission’s rules for political advertising just don’t order enough transparency. The FEC requirements for labeling transactions, for instance, are too loose; an ad purchased on Facebook could be logged as “digital ads” or merely “ads.” More alarmingly, as much as 94 percent of ad spending may go through consultancies, but campaigns don’t have to report how those firms spend money on their behalf.
然而，即使在今天，这些问题的答案也很难得到。联邦选举委员会关于竞选广告的规定没有要求足够的透明度。例如，联邦选举委员会对标记交易的要求过于宽松；在Facebook上购买的广告可能会被记录为“网络广告”或仅仅是“广告”。更令人担忧的是，多达94%的广告支出可能是通过咨询公司进行的，但竞选活动不必报告这些公司是如何代表竞选人花钱的。
The flaw may enable the laundering of illegal spending, but it also means it’s impossible to tell how campaigns adjusted ad buys in response to ad bans. “We can see money going in,” say the Duke authors, “but we can’t see money going out.” Without knowing who shifted what amount of spending where, there’s no way of understanding whether bans affected office-holders more than challengers, or Democrats more than Republicans. There’s also no way of knowing how much speech was actually suppressed, or how much was redirected to forums with more robust fact-checking rules. Mandated and standardized ad archives would help, too, especially if they included the firms that coordinate ads across thousands of websites but don’t host them themselves.
这个漏洞可能会使非法支出洗钱成为可能，但这也意味着不可能知道竞选活动如何调整广告购买以应对广告禁令。“我们可以看到资金流入，”杜克大学的作者说，“但我们看不到资金流出。”如果不知道谁把多少支出转移到了哪里，就无法了解禁令对在职者的影响是否大于挑战者，对民主党的影响是否大于共和党。也没有办法知道有多少言论实际上被压制了，或者有多少言论被重定向到有更健全事实核查规则的论坛。法定和标准化的广告档案也会有帮助，特别是如果档案内容包括那些在数千个网站上协调广告但自己不托管广告的公司。
These changes will likely require a mix of revisions by the FEC and legislation. These modest improvements to policy could have a profound impact on elections to come, at the least producing meaningful interventions against misinformation next time rather than only sound and fury.
这些变化可能需要联邦选举委员会和立法机构的共同修订。这些对政策的温和改进可能会对未来的选举产生深远影响，至少会在下次选举中对虚假信息进行有意义的干预，而不仅仅是喧嚣和愤怒。