Manufacturing the News

Polls paid for by media become news stories

This is my 27th letter to the Washington Post. I generally do believe that the letters I write have a chance of being published, but I knew this one was a bigger stretch than usual. The letter accuses the Post of paying for polls on topical issues, then managing the way the results are released in their paper. That's actually not where the stretch lies; it's what they do. But it is also easy for editors and reporters to argue that the poll results they report actually examine public feeling to a depth that readers cannot. Still, polls generally report who is hot and who is not. Much of the additional detail organized in reports of polls as if it were factual is actually subject to both interpretation and statistical error.

My guess is that newspapers started polling as a defensive move; an attempt to protect their stories from the spin that owners of private poll results might introduce. After all, politicians were polling before newspapers, and leaking the results of polls to targeted journalists in an effort to influence coverage of elections. That doesn't happen so much now that everybody, politicians and journalists come armed with poll results. Campaigns don't try as hard to use their polls to spin stories, primarily because they more often use their polls to identify target audiences and develop messages aimed at groups of voters. Nor would spinning work as well in an environment where a newspaper can report their own, perhaps different, poll results. But the original defensive purpose for polling has changed--survey results are now news, and journalists no longer have to hit the streets to get the story.

Editor,

A quick look at the last two issues of the Post provides a clear example of how print media manages the news, rather than simply reporting it. The front-page over-the-fold story on Sunday, Jan. 31 (“Fenty’s approval ratings plummet”) says that a “new Washington Post poll” shows that DC Mayor Adrian Fenty’s popularity among residents has dropped dramatically since a 2008 poll.

A day later, in a metro story (“D.C. Schools Chancellor Rhee’s approval rating in deep slide”), we read that the same “new” poll reveals that the public perception of Chancellor Rhee is also much more negative now than it was two years ago. But the serial nature of the stories raises the question of exactly what journalistic criteria might have dictated reporting the two stories separately. Personally, I know of none. But I can imagine the business considerations that went into deciding that the stories would have more impact, if they ran separately on Sunday and Monday rather then running together on, say, a low-circulation Saturday.

Speculating in this manner raises further questions, I think. For example, if poll results are news, at all, then why wouldn’t the Post report the results when they first become available? But, perhaps more to the point, why would the Post bother to write about poll results, at all? Editors and reporters routinely assert that journalists investigate and report stories; they do not manufacture them. But news stories based on polls that newspapers pay for seem perilously close to product manufacturing.

Perhaps editors assume that none of us actually talk to each other. It is therefore a great service of the Fourth Estate to share the news with me that my neighbor to the east likes Fenty, while my neighbor to the west does not. If that is the assumption that editors are making, it may be helpful for me to share this fact: by and large, we do know what our neighbors are thinking.