Brendan Nyhan cautions journalists about their role in the Iowa caucuses:
Unfortunately, the “meaning” of the caucus results is not always clear. These rough edges are typically sanded away in post-Iowa reporting and commentary,  which tends to emphasize the order of the finish (even when the margins between candidates are small) as well as unexpectedly weak or strong results.
….The result is a refraction effect in which journalists help make Iowa influential and then report on its “effects” without acknowledging their role in the process or the often arbitrary nature of the distinctions that are made among the candidates. This is a recurring problem—the norms of journalism demand that reporters exclude themselves from the stories they write, creating a troubling lack of self-consciousness about their own role in the process….These incentives are especially problematic in campaigns since journalists have a strong rooting interest in continued conflict and dramatic storylines.
For what it’s worth, the Republican race has been so chaotic this year that I think the media’s role has been less than usual. In fact, I’m hard put to give the press much credit for any of the ups and downs we’ve seen over the past few months, largely because they’ve simply happened too fast. They’ve been more the result of tea party litmus tests, serious gaffes from the candidates, and immense amounts of super-PAC spending.
As for reporters most likely emphasizing the order of finish after tonight’s caucuses — well, what else would they do? Close or not, the order of finish is what everyone cares about. The media can hardly be blamed for that.
For the most part, I suspect the media reflects more than it refracts. That is, it mostly operates as a pretty reliable mirror of what party elites and Beltway thought leaders believe. This year, you can add intra-party feuding to that list. You may or may not think this is a good thing, but overall I think there’s been less independent role for reporters this year than usual. Candidates just don’t need the press as much as they used to, and that’s shrunk both the media’s role and its influence.
And one last thing: the idea that reporters are “rooting” for conflict and dramatic storylines is so common that it’s practically conventional wisdom. It’s true, of course, that reporters are drawn to news, which is frequently rooted in conflict of some kind. But are they actively hyping otherwise banal events because they don’t want the campaign to end? That’s what people mean when they say this. And it sounds good! But is there any actual rigorous evidence for it? I’ve never seen any.