On Thursday, Todd devoted a segment of The Daily Rundown, his morning show on MSNBC, to talking about the poll with McInturff and fellow pollster Fred Yang. After a modest, low-key introduction (the poll was “the big story, frankly, driving the day”), Todd turned it over to McInturff, who said:

The most important point is how corrosive the economy is. After just a little bit of feeling better and April and May, once again, people dropped in economic confidence. Only 33 percent say they think the economy will get better. And the most powerful number is that 82 percent of Americans are dissatisfied with the country’s economy, half very strongly. No American president can stay in office and have those kind of economic numbers, and that kind of dissatisfaction, on top of—and you add the oil spill, which is a two-month, how-do-we-fix-this story, and not presume sooner or later you’re going to see a push-down on your job approval.

Yes, he said the words “oil spill.” He also said the words, “The most important point is how corrosive the economy is.” And, “the most powerful number is that 82 percent of Americans are dissatisfied with the country’s economy.” Pretty clear, right?

Apparently not. Todd, relentlessly on message, followed up by turning to Yang and asking: “Is it fair to say that the oil spill is the difference here?” That drew this ambiguous response:

I think it’s fair to say that. I think it’s not the oil spill per se—just two more months, as Bill said, of just bad news.

At the end of the segment, after sneaking in another remark about the “economic weight” dragging down the president, McInturff made a game effort to clear this up:

This is the Gulf spill in a second. The president’s been stopped—he’s had two months where he’s not been able to talk about jobs, jobs, jobs. And it’s pushed [aside] any ability the Democrats and the president have to try to reframe an economic message to try to get people feeling better. And I think that’s a lost opportunity.

There are reasons to doubt that with better “messaging” the president would be able to persuade voters that a lousy economy is not, in fact, a lousy economy. But the broader point, made clearly by McInturff, is that while the oil spill isn’t helping, the economy is the key factor shaping public opinion. It takes a perverse, willful adherence to your preconceived narrative to get this kind of input from the expert you are paying for his expertise and render it as “Poll: Spill drags the president’s ratings down.” (The Journal’s coverage was much better on this score, though it still didn’t take into account other polls.)

Applied properly, polls can tell us something about what the public thinks—discrediting bogus storylines and beating back speculation from pundits who “know” the public agrees with them. But when you hear someone claiming that a single poll has identified an “inflection point,” or the precise moment when a specific issue has “broken the camel’s back,” you’re hearing a claim for something polling can’t do. Time to tune that person out, and start paying attention to someone who’s interested in what the numbers can tell us.

P.S. Pollster.com contributor Brendan Nyhan makes a similar point at his blog.

If you'd like to get email from CJR writers and editors, add your email address to our newsletter roll and we'll be in touch.

Greg Marx is an associate editor at CJR. Follow him on Twitter @gregamarx.