The Daily Mail has been in trouble again recently over its policy of how to moderate user comments on its site. Malcolm Coles, writing for The Media Blog, reports that despite (in this case) user comments supposedly being moderated before going live on the site, extremist racist comments got through the moderation system, and despite being reported for abuse (by MailWatch and the Media Blog in this instance), were left on the Mail’s site from 6.30pm to 11am the following morning, before being taken down. MailWatch has the details, including the comments that were allowed through.

The Mail changed its moderation policy earlier this year (its policy can be viewed here), and now selects when to moderate comments (whether and how comments are moderated or not is made clear under each article). In unmoderated sections, the onus is on readers to report abuse. Judging by the example cited by Coles, this policy isn’t working, as racist or abusive comments are clearly getting through the system, being stopped by neither pre- nor reactive moderation.

There’s another interesting twist to the Mail’s comment system: readers can vote for their favourite comments, and can view comments by the most ‘popular’ (i.e. the most voted-for). Many of the people who care enough to vote on other people’s comments will be those who have passionate, or extreme views on an issue. Racists, bullies, political extremists all fall into this category – as is made very clear by the example above. This has the (intended?) effect of bringing the most reactionary comments to the top of the pile, and stoking the fires of the debate.

It seems to me that if you are going to allow your readers not just to comment on articles, but to vote on those comments, you must have a moderation policy that is consistent and effective, to prevent abusive (in this case, racist) comments from becoming the ‘favourites’ on a site. The site owner will be associated with that racism. Ethics aside, on a purely commercial basis this association could harm advertising revenues – how many mainstream advertisers will want to be associated with racism on a news site?

These days, news sites are not just about delivering news that has been researched by journalists. They are interactive communities, where readers can join the discussion. The media owner has a moral responsibility (if not a legal one) for content on that site, whether it is uploaded by its journalists or its readers – all of whom publish within that community.

There is also evidence that users prefer to comment in a moderated environment. A study in 2006 from the University of Missouri-Colombia’s school of journalism found that users preferred to comment in a moderated environment . Interestingly, this was not predominantly because users were concerned that they’d be exposed to abusive messages, but because they didn’t want a discussion to get sidetracked (by spammers, or thread ‘hijackers’, for example). The study found that moderating news comment sites actually helped to encourage more people to participate.

All the major news sites have strict user guidelines to attempt to control user behaviour and comments. Most sites post moderate (i.e. check that comments comply with site rules after they have gone up on the site) and reserve the right to delete anything that breaks the rules. The Independent, for example, makes it very clear that it will ban users who break these rules from participating on the site again. All the major news sites state that they accept no liability for comments that break guidelines; the FT explicitly makes users financially responsible for any claim resulting from a breach of site rules.

But these rules are meaningless if they are not enforced. What the vociferous minority of your readers think is acceptable, and what you and your advertisers think is acceptable may be worlds apart. Whilst traffic is undoubtedly up, the Mail may yet rue its decision to change its moderation practices.