Skip to main content

Hillary/Trump and political journalism

This is a blog, about politics, in 2016, so we're long overdue for something about Hillary Clinton and the media. I have started to get a bit disoriented about what the point of a lot of political journalism being written this year is.

Contrast two different types of writing about politics. (There are obviously other sorts.) On the one hand, people write about something - a policy, a candidate, a statement - is bad. This can come in a lot of different flavours and run a spectrum from quite even-handed and analytical to a full-blown opinion piece. But the core uniting feature is that the article (column, post, whatever) ends up enunciating something which the author themselves might take as a reason for voting one way or the other. On the other hand, you have horse-race analysis, where the writer asks who's winning, how a candidates' actions might play with the public, and so on. In theory, these pieces are united by being the kind of think a politician might be told by their advisors. Outside very particular contexts, it wouldn't make much sense to read this kind of article as giving you a reason to vote for or against a candidate: in general, the fact that a candidate is gaffe-prone or unpopular in Florida has no bearing on whether you should think they're the best person for the job. In practice, though, negative stories of the second kind do make people less likely to support a candidate, and the boundary between the two types of piece isn't particularly clear.


Last week Hillary Clinton was sick with some form of pneumonia, which she didn't disclose to anyone until after she'd been filmed almost collapsing whilst walking. (I have not actually watched the infamous video, so this may not be a great description of it.) Now almost no journalists go in for the line that having pneumonia disqualifies you from being President, but we instead got a spate of articles and discussion about how she should have publicised her sickness earlier. Not doing so, they argue, has worsened one of her gravest problems - that people perceive her as dishonest and untransparent.

I may just not have a feel for this criticism, but I really can't see how not having disclosed her pneumonia is a sign that should make us worry about Clinton's merits for the job of being President. Maybe you want a President who's more generally up-front and discloses lots of information as a matter of course, but it's honestly a pretty weak complaint. That is: you can't write a piece of the first type about this issue without it being pretty bad.

But you can easily write the second type of article, which is why there's been so much focus on Clinton's existing 'honesty problem' and how this exacerbates it. This is where I start to get a bit dizzy. Almost nobody follows political news very closely, and certainly not as closely as political journalists. Many voters think Hillary Clinton is dishonest, but they're extremely unlikely to know the timeline of when she got sick and when she made her various changing explanations. The vast majority of people would never connect the pneumonia disclosure to their general feeling that Clinton is dishonest - if it weren't for the rush of articles and TV talking heads making the connection.

Mostly, that commentary wasn't saying that this incident should make people distrust Clinton and be wary of electing her. It was only saying it would have that effect. (Type two, not type one.) But this kind of horse-race-analytics prediction helps to create the result its foreseeing. And while that's great for making your predictions come out right, the upshot is that journalists' coverage of the Clinton/pneumonia/disclosure issue ends up increasing doubts about Clinton's fitness for the presidency, even though almost none of the journalists think it actually should.

There's something deeply dubious about making predictions about voters' responses when, because most audiences don't parse commentary closely enough to sort it into 'predictions' and 'advice' boxes, those predictions are one of the main factors in determining voters' responses. That dubiousness - obviously! - doesn't go away if your predictions end up being validated by polling. I think there's a very good case for saying that people should refrain from horse-race commentary except when it's backwards-looking - 'two weeks later, polls suggest that this incident raised voters' distrust of Hillary Clinton' - to avoid this kind of perverse moulding of public opinion.

Concluding note: I'm seriously worried about how the presidential debates are going to be covered. The reality is that we can confidently say, a week in advance, that Clinton will win the debate convincingly. She will be more measured, she will tell far fewer lies, she will answer questions more directly, and she will have a vastly better understanding of all the issues that are raised. Can you just write that? Maybe not, especially given you could have written it a week ahead. It's too easy to imagine the commentary - Trump's brash style is very effective, voters are going to like the way he went on the attack, they won't like how cautious Clinton seems - and then...