It has become a ritual of each campaign cycle. News media critics lament the prominence of poll driven, "horse-race" campaign coverage and urge a greater focus on "the issues," yet horse-race polling and coverage continue to proliferate. But rather than an endless cycle of hand-wringing that leads nowhere, perhaps we could just do a better job reporting the horse race.
Last weekend, for example, Washington Post ombudsman Deborah Howell reported that horse-race stories outnumbered those about issues during this campaign by more than 3 to 1. "Horse-race stories can be fascinating," she wrote, "but hard nosed journalistic comparisons of candidates' stands on the issues give readers what they need to know."
Similarly, New York Times public editor Clark Hoyt recently commended his colleagues for standing "almost alone among major news organizations" in having "played down polls" in their coverage leading up to the February 5 primaries.
His evidence? Hoyt noted that on the Sunday and Monday before Super Tuesday, both the Washington Post and USA Today reported their own national polls "at the top" of their respective front pages while the Wall Street Journal gave front-page play to results from three prominent polls. Hoyt did not mention it, but even the Huffington Post, which had recently relegated polls to a "HuffPollstrology" page that also includes candidate horoscopes and primary state weather forecasts, succumbed with a huge front page headline linking to a new CNN poll ("DRAMATIC SHIFT, GOP ALL BUT OVER").
The Times, on the other hand, opted against conducting a national poll just prior to Super Tuesday. According to Hoyt, it also avoided any mention of polls in its front-page Sunday story on candidate advertising strategies and relegated mention of "poll numbers" to paragraph 17 of its front-page story on Monday.
Sounds impressive. Of course, readers of that Sunday's Times needed only turn a few pages to find a 1,300 word story, which reported in the second sentence that a "variety of weekend polls showed Senator of New York was locked in a tight race with Senator of Illinois for the Democratic nomination." The next sentence provided the horse-race numbers from the latest poll conducted independently by CBS News, ordinarily the Times' polling partner.
The Monday story that, as Hoyt conceded, "conveyed the impression that McCain was ahead and the Democratic race was tight," did so with the following words in its lead paragraph: "Senator , buoyed by new polls and endorsements, appeared in an increasingly commanding position on Sunday." It went on to report that Clinton and Obama "were enmeshed in a tough national fight, illustrated by polls showing the race had tightened both nationally and in key states."
If the New York Times "played down" polls before Super Tuesday as compared to other prominent news sources, the distinction was subtle at best.
The underlying problem here is that news consumers seek out this sort of horse-race coverage because it interests them far more than dry summaries of candidate issue positions. Eight years ago, three political scientists conducted an experiment that sent a computer CD full of campaign stories to a nationally representative sample of Americans with computers. The CD was set up to record which stories they clicked on and read. They found that stories that readers sought out most "were those focusing on the strategy-horserace aspects of the campaign," while "news reports on the issues and nomination process were altogether ignored."
On Super Tuesday, McCain prevailed while the Democratic contest was close, so the national polls gave a reasonably accurate view of the horse race. Within individual contests, however, polling data varied considerably. In the most infamous example, the results from final polls in California varied widely, showing anything from a 10-point Clinton lead to a 14-point Obama advantage (Clinton won by 9). Four other states showed polling gaps nearly as wide, leaving observers scratching their heads and questioning the value of pre-election polling.
How can we do better? We can start with polling and reporting that does a better job depicting the horse race. Yes, primary polling is challenging, and the hard-to-gauge turnouts and uncertain preferences of this year's presidential contest make it exceptionally tough. All too often, however, the most respected public pollsters abandon high-profile statewide contests as the election nears. "We tend to stand back," the Times' Janet Elder told Hoyt, because "opinion is in such flux in the last days before people vote."
But if opinion shifts are decisive to election outcomes, and if news organizations want to better understand their basis, why not design longer analytical surveys that help monitor and explain the basis of those final shifts as they occur (rather than just passing along second-hand accounts of "polls" that track only the horse-race question)?
Second, we ought to use the drama of the horse race to draw readers into coverage that connects campaign strategies to the underlying contrasts (on issues, qualifications, leadership styles) between the candidates. If a story attracts readers or viewers interested in "who is going to win," how well does that story highlight the debate between the candidates? How well does it use the tools of its particular medium (hyperlinks, sidebars or on-air references to Web site URLs) to promote stories or resources that give uncertain voters "what they need to know" to make better decisions?