A majority of Americans distrust the media. Scholarly assessments of the media are usually no better. And these sentiments often reach fever pitch during a political campaign when the news media are accused of emphasizing trivia or being flat-out biased toward one candidate.
In “The Gamble,” Lynn Vavreck’s and my new book on the 2012 presidential campaign, we have a lot to say about what, and how, the media did in covering the campaign. This is one unique feature about our book, compared to the campaign books written by journalists. We make the media a central character — an actor in the process, not just an observer. We do this with data on 11,000 different news outlets gathered by the company General Sentiment, which gauged not only how often the candidates were mentioned but how positive or negative that coverage was. Here is what we found:
1. In the Republican presidential primary, news coverage drove the candidates’ surges in the polls.
The ups and downs of the primary candidates were not “random.” They followed a consistent pattern. A candidate who hadn’t previously received much coverage would do something that was judged noteworthy. For Rick Perry, it was simply getting into the race. For Herman Cain, it was a victory in the Florida straw poll. For Newt Gingrich, it was the debates before the South Carolina primary. For Rick Santorum, it was first his victory in the Iowa caucus and then his later victories in Colorado, Minnesota and Missouri.
In the wake of events like these, positive news coverage of the candidate shot up — a process we call “discovery.” The poll numbers followed suit, as the statistical analyses in the book suggest. We argue that this pattern is understandable — the news covers what is new, after all — but sometimes problematic. The Florida straw poll was a meaningless event — a “pseudo-event” — and yet it not only earned Cain coverage but often led media outlets to frame that coverage in a specific way: as a Cain victory over Perry. “Cain Upsets Perry at Florida Straw Poll,” declared USA Today. “Cain Upsets Perry in Florida Republican Straw Poll,” declared Reuters. “Herman Cain Upsets Gov. Rick Perry to Win Florida GOP Straw Poll,” declared Fox News. While journalists should report on the day’s events, the drive for newsworthy headlines sometimes elevated events whose significance was arguable at best. (The same would be true with the focus on “gaffes” during the general election.)
2. In the primary, news coverage helped end these surges as well.
After their discovery, we show that front-running GOP candidates soon experienced significant scrutiny. The media began to investigate their records, and this usually led to more negative news coverage — something that the opponents of the candidate being scrutinized were only to happy to facilitate. One example of this was Politico’s story about the sexual harassment accusations against Herman Cain. This scrutiny reflects positively on the news media because such coverage helps voters hold politicians accountable for their past deeds. In Cain’s case, the effect of this scrutiny on voters’ opinions of him was clear. Below is a graph from the book. It shows the percentage of coverage of the Republican candidates that focused on Cain (gray line), that percentage multiplied by a measure of how positive or negative the coverage was (black line), and Cain’s national poll numbers (the dots). The scrutiny is reflected in the shifting tone of coverage of Cain — the black line moving downward — even before the Politico story broke.
3. In the general election campaign, it was the other way around: the polls drove the news.
This isn’t surprising. The polls in the general election campaign moved very little — mainly after the Democratic convention and the first debate. Who “won” the news cycle simply didn’t matter that much. That poll numbers drove news coverage is also unsurprising. When news coverage focuses on the horse race, as it did in 2012, the candidate leading that race tends to look better in news coverage. Meanwhile, the trailing candidate is treated to more stories about what is allegedly wrong with their campaign — like this one about the Romney campaign when it was trailing in mid-September.
4. Overall, media coverage of Obama and Romney was actually fair and balanced. No, really.
This is the sort of statement that tends to make partisans on both sides irritable. (Because of this.) But when we examined General Sentiment’s measure of how positive or negative coverage of Obama and Romney was, neither candidate had a chronic advantage. Here’s our graph from the fall campaign:
You can see predictable spikes in the volume of coverage (the gray line) around big events. You can also see that Romney received somewhat more coverage than Obama in the outlets that General Sentiment was monitoring. But most importantly, you can see that once we factored in the tone of the coverage — the black line — the media favored neither candidate consistently. Sure, the week after the release of the 47 percent video was a rough one for Romney. But September wasn’t that great for Obama, either. Similarly, coverage of Obama was more negative, and coverage of Romney more positive, after the first debate — as the Project for Excellence in Journalism also showed using a different methodology — but this didn’t last long.
Ultimately, when we looked at the average across the entire fall campaign (and the same was true in the summer), we found that the tone of the coverage of the two candidates was almost exactly the same. Neither was covered much more positively or negatively than the other. This is consistent with the Project for Excellence in Journalism’s research and with scholarly research on previous presidential elections. Taken together, this is comforting evidence that the media writ large can approach election campaigns with minimal partisan or ideological bias.
5. The news media are more prone to “root for the story” than “root for the candidate.”
Given the apparent lack of partisan bias, it is clear that the media was not rooting for Obama or Romney. But at times they were rooting for something else: a good story. This was most visible during the third week in October, when there was a spate of stories, beginning in Politico, about Romney’s alleged “momentum.” There was nothing in the polls to justify this notion (see here or here). So what did justify it? When another media organization published a story suggesting that Romney had “momentum,” I wrote to one of the reporters whose byline was on the story. I pointed out that there was no evidence in the polls and asked this reporter why that word was included in the story. As we write in “The Gamble”:
Indeed, when asked whether Romney truly had momentum, one reporter for a major newspaper admitted to us that the story was really about what other reporters were saying, not about any actual change in the polls. This reporter said, “What we call momentum is more like narrative, and we’re buying into that.”
That’s a pretty telling comment, and it illustrates one of the dangers of reporting on a campaign — and one of the reasons we wrote our book. The media should want to write interesting stories, but there is always the risk that meaningless events or non-existent trends get puffed up for no reason. In the book, we argue that there is a way to report the events of the day without exaggerating the significance of those events:
Another goal of journalists is to be skeptical about what the candidates and their campaigns say. Campaigns are forever pushing their own spin on events, and journalists know better than to believe it. This produces any number of news stories that investigate, fact-check, and critique the candidates’ claims. So when candidates or their spokespersons go around promoting some moment as a potential game-changer, journalists can draw on moneyball — from political science, poll numbers, history, and so on — to identify that as spin, too. For example, the Romney campaign was free to go around claiming that they had momentum late in the race. Indeed, since they were behind at that point, it made sense for them to claim to be gaining ground. But there was no reason for reporters to imply that this was true when the polls had not moved for three weeks. The lesson is simple. Journalists have to report what the campaigns are saying and doing on any given day — that is their job — but when journalists evaluate what the campaigns are saying and doing, they can use social science and a closer look at the available data to separate truth from spin.