Now that the first part of the election cycle is winding down, journalists are taking a short breather between the primaries and the presidential campaign itself.
Along with catching a breath, we can also take time to evaluate our coverage so far. For example, have we used polls responsibly or has our over-reliance on these props negatively affected the recent coverage of the Democratic primaries?
The media has made some mistakes, says Mark Jurkowitz, media critic for the Boston Globe.
“The punditry predictions about Howard Dean being the front-runner and that he might be almost unbeatable – which were based on a number of polls – were certainly way wrong, as we found out,” he said.
But the polls weren’t all to blame, he added. “In the last few days, the polls did show movement away from Dean and some dramatic movements towards Kerry.”
Jurkowitz says that many political reporters frequently fall back to a “polling horse-race” type of coverage, especially during a multicandidate primary season.
“Whether accurate or not, they create some sort of empirical sense of what’s going on in the race,” he said. “And there’s some psychological relief when reporters can look at some numbers, something solid as a snapshot of the race.”
BRING ON THE BEAT REPORTERS
But there is another important reason why political coverage is often so dominated by poll numbers, says Jurkowitz. And that’s because, too often, it’s the political reporters who cover politics.
Now that might seem to be self-evident, but Jurkowitz points out that the primaries are about more than just politics. They’re also about issues such as education, the environment and international affairs. It would make sense to get other reporters involved in covering the elections, he says.
“If we’re going to talk about taxes, why wouldn’t business reporters be the best people to cover that aspect of the race?” he asked. “If we’re going to talk about Iraq, why wouldn’t Pentagon or defense reporters be the best to cover that part of the race?”
Political reporters, by comparison, often are not the best experts on the issues that a news organization might have.
Newspapers are better at this than other media outlets, he added.
“I think print does a better job of more sophisticated coverage than television does,” he said. “If you look at major newspapers, you don’t see stories dominated by poll numbers every single day. You get a lot of different stories.”
Television, by comparison, tends more toward shorter and choppier stories, or the kind of insider-baseball analysis that can fill long hours on 24-hour cable news stations.
“Political reporters understand the language of strategy and horse race better than anything else, so they tend to gravitate to those kinds of stories and to gravitate to polls,” he said. “And frankly, it’s easier to cover polls than to try to get a handle on the tough issues.”
As a result, he said, political coverage often revolves around an overuse of polls.
“We in the media pay attention to these polls, and the insiders pay attention to these polls, but the truth is the public is not nearly as interested in those kinds of things,” he said. “They’re certainly not as interested in those kinds of inside-baseball strategy things as the media.”
Polls do have a value, he added.
“I personally am interested in them, when I see them, I look at them, and they have broadly tended to be accurate over time,” he said. “Exit polls are pretty good because there’s interesting information there about why people voted the way they did. I just think people should be more discriminating about how they use them.”
DON’T TRUST THE EARLY POLLS
In particular, reporters should be more discriminating about how they use the earliest polls of a campaign, before the public has a chance to find out who the candidates really are.
According to Steve Simurda, a journalism professor at the University of Massachusetts at Amherst, early polls have the highest error rates but are the most significant when it comes to fund raising – and on determining coverage.
“We start out with eight or nine candidates in the primaries,” said Simurda. “There is no news organization that is going to place reporters full time on all of those campaigns, and you could argue that nor should they be expected to. So they have to make a choice. Who are we going to cover? Who deserves to be covered part time? Who deserves to be covered more? Who deserves to be covered less?”
Too often, the decisions are made on the basis of polling numbers, even though the polls have very little to say about which candidates resonate best with the American public.
“That becomes a self-fulfilling prophecy,” Simurda said. “I’m sympathetic with the candidates who say, ‘You don’t cover me because I’m not a major contender, but I could become a major contender if I got more coverage in the media.’ It’s a Catch-22 situation.”
So the early coverage focused on Howard Dean and John Kerry.
“At no point were most news organizations giving more than two or three candidates full-time reporters assigned to their campaign,” Simurda said. “So you had two or three people getting regular coverage and the rest of the candidates getting the occasional story. And if one person rises above, like Wes Clark rose a little in the polls, they’ll put somebody on them.”
Simurda says he doesn’t know if there’s a solution to this problem.
“I would like to see less reliance on polls in general,” he said. “And, at least at the beginning of a campaign, the press has to be vigilant to provide an opportunity for all the candidates to get a chance to get their message out to the public. Look at this campaign – it took forever for John Edwards’ campaign to take off. You can’t always predict these things. If Edwards had more coverage early in the campaign, could he have provided a more formidable challenge to Kerry? We’ll never know.”
News organizations do occasionally examine the intrinsic value of political candidates, Simurda admits.
“The candidates meet with the editorial boards of major papers in the states where they had campaigns,” he said. “But that’s done more for endorsements than for determining how to cover a campaign.”
Polls are least accurate at the start of a campaign, agrees Bruce Altschuler, professor and chair of the political science department at SUNY Oswego. Altschuler also is the author of the book “Keeping a Finger on the Public Pulse: Private Polling and Presidential Elections.”
“A poll taken six months before the first primary is going to tell you that a candidate or two has name recognition and not a heck lot more than that,” he said. “In Iowa, where most people don’t vote in the caucuses, polls tend to be extremely unreliable, because it’s hard to predict who’s going to turn out.”
It doesn’t only affect coverage and fund raising but voting patterns as well, he said, when the polls get more play than they deserve.
“A lot of voters are going to engage in strategic voting,” he said. “They’re not going to vote for a candidate who has no chance, and they might instead vote for their second choice.”
But Altschuler sees the system as an inevitable outgrowth of our democratic process and our market-driven media decisions.
“Especially on television, the media feels that the voters may have a short attention span and discussing policy issues is complex and not overly visual,” he said. “So we cover it the same way we cover sports. We’ve got the baseball scores, and we’ve got the political scores.”
It’s true that reporters can take a more substantive approach to political coverage, but it might not do much good, he says.
“If the public really wants substance and demands it, then it will watch those programs that have it,” he said. “But the TV networks will tell you that when they try to put on substance, the ratings go down and their profits are lower. Unless we, as the public and as potential voters, watch on those occasions when there is substance – until that happens, things aren’t going to change.”
It doesn’t help that controversy and scandal also crowd out coverage of legitimate, substantive issues.
“Is the public dissatisfied with the kind of coverage it receives? Apparently not, judging by the ratings,” he adds. “And in a market-driven media, where there is no economic reason for it to change, it won’t change.”
DON’T SKATE ON THE MARGINS
If reporters are going to use and overuse polls, then the least they could do is use them accurately. According to Stephanie Larson, professor of political science at Dickinson College, all of the problems associated with the overuse of polls are exacerbated when journalists make basic mistakes in how polls actually work.
Larson recently published an article in the Harvard International Journal of Press and Politics on the misuse of polls on national television news during the 2000 general presidential election.
Then, as now, reporters kept making the same mistake: They undervalued the margin of error.
There were other mistakes, but this was the most egregious since misunderstanding the margin of error caused reporters to proclaim losers and winners when in fact there was a statistical tie (see sidebar).
If the margin of error is three points, then candidates have to be at least six points apart in order for one to have an actual lead over another, instead of being in a statistical tie, said Larson.
“That’s one of the ways reporters do misjustices to the polls, by building whole explanations about who’s ahead and who’s behind,” she said. “They were doing that as well here in 2004 when someone had pulled ahead, but maybe they had and maybe they hadn’t.”
POLLS CAN’T SEE THE FUTURE
Another mistake reporters make is thinking that polls say something about the future, Larson says. They don’t.
When reporters make the mistake of thinking that polls are about anything but what a potential voter is thinking at a particular moment, they start making guesses about the general direction of the election, she says. As a result, it looked as though they were wrong a lot this past election.
“There is nothing wrong with voters mulling it over during a campaign,” she said. “That’s what a campaign is supposed to be for. It isn’t unreasonable for a voter to think, ‘I’m leaning towards Edwards,’ and then a week later to say, ‘No, I’m going for Dean.’ Back when it looked as though the polls were off a lot, in part that was because reporters used the polls to say what would happen in a week instead of saying how people feel this week.”
This is especially true early on in a campaign, when voters are still learning about the candidates. Later on, after the primaries, when the race for president is down to two or three people, then voters are less likely to switch allegiances on a regular basis.
“In the general election, people really have made up their minds,” she said. “But in the primaries, the decision-making is more difficult, and a poll is inherently going to be less reliable two weeks before a primary than on primary day. And, too often, any change in the poll – while being kind of exciting to reporters and kind of interesting – is used to demean voters. That is a mistake.”
MORE POLLS, MORE, MORE!
Another mistake reporters make is not to use the polls enough, Larson added.
“The attraction of polls is pretty much one or two questions – who’s ahead and who’s behind,” she said.
But polls can do more than that. Some polls get more elaborate and ask voters how angry they are at potential candidates, for example.
“But those don’t get talked about as much, and it still doesn’t exhaust the kind of information you can have in a poll,” she said. “The poll is an opportunity for the citizens to be part of the dialogue. Elections are about citizens, and we forget about that. It’s the public’s opportunity to speak. All we usually hear about is who they voted for. And with polls, we can hear a lot more. In polls, we can ask the public what they think, what they want, what they know, but that’s now how reporters think about it.”
The reason? Habit, Larson says. “When I analyze the election news, it doesn’t look very different from election cycle to election cycle. You get a lot of the same patterns of coverage. There’s always going to be a story of how New Hampshire is getting crowded with too many journalists. That story shows up every single election cycle – it’s a habit.”
Similarly, she said, reporters get into a certain habit of reporting on poll numbers.
“Reporters are attracted to news that’s new and interesting and changes,” she said. “To the extent that polls change, that a week ago Kerry wasn’t as popular as he is today, that fits into what journalists want to see.”
So the lesson of this year’s primary season? Pay less attention to early polls and more attention to margins of error, and look for polls that do more than just identify the front-runner.
And, last but not least, give your political reporters a break and have someone else do some campaign coverage as a change of pace.
Your audience might thank you.
Maria Trombly is a freelance writer and co-chair of SPJ’s International Journalism Committee.