Home | Opinion | Columnists | Mark Mellman

Pick a story, any story

I lived in New York City during the heyday of Three-Card Monte. Walking down any street, at any time, you could not avoid shouts of “Pick a card, any card.”

Today’s political discussions evoke similar feelings. Everyone has a theory, a story to tell, and the proliferation of primary polls provides everyone with evidence “proving” his or her interpretation.

Say you want to argue that the controversy over Obama’s “bitter” comments had no impact on the Pennsylvania primary — you’ve got the data to back up that claim. On April 8, days before the comments, the InsiderAdvantage poll showed Hillary Clinton with a 10-point lead. On April 20, a week after the controversy erupted, Clinton’s lead was exactly the same 10 points. No impact.

Public Policy Polling might even allow you to argue that Obama’s sociological analysis helped him. Just before the comments, Clinton clung to a three-point lead, in their telling. Within days of the comments, Obama had turned it around and led by three.

However, let’s say you wanted to press the opposite claim — that Obama’s comments dealt a devastating blow to his Pennsylvania prospects. Just turn to ARG. The week before the comments surfaced, Clinton and Obama were tied, they say, but as the story broke she turned that into a 20-point blowout, which settled back to a 10-point victory.

So we have same-poll to same-poll comparisons telling us the “bitter” comments helped Obama and hurt him — and that they had no effect. Pick your story and you have the data to back it up.

What of the Rev. Wright controversy? In a story headlined “Polls show Barack Obama damaged by link to Reverend Jeremiah Wright,” the Times of London “proved” their contention by citing Gallup — “A new national Gallup tracking poll shows Hillary Clinton regaining her lead over Mr. Obama for the first time in a month, now leading 49 percent to 42 percent, a 13-point shift to the former First Lady in less than a fortnight.”

Under a different headline — “Wright Controversy Doesn’t Hurt Obama, Poll Shows” — The Wall Street Journal argued the opposite case from a different set of polls: “The [Pew] survey … still puts the Illinois senator in the lead for the nomination, 49 percent-39 percent, largely unchanged from the 49 percent-40 percent lead Obama held a month earlier.”

What role did Hillary Clinton’s now-famous “3 a.m.” ad play? The day before it was unveiled in Texas, ARG claimed Obama held a seven-point lead there. Less than four days later, the same pollster reported a dramatic turnaround, with Clinton going into Election Day three points ahead, apparently thanks to the ad.

Not so fast, though. Public Strategies said no impact. Clinton was one point ahead both before and after the ad. InsiderAdvantage had Clinton four points ahead before the ad and five points ahead after.

IVR Polls hints at a slight negative impact from the ad. It narrowed a five-point Clinton lead to three points.

Opposite readings of the general election also now abound. Clinton leads McCain by nine points in the latest Associated Press poll, but is two points behind in a Rasmussen survey done the same days. Obama is ahead of McCain by two in the AP poll, but even in Rasmussen. So the AP poll purports to tell us that Clinton is stronger, while Rasmussen awards Obama that title.

Analysts often tell somewhat different stories about what really happened in particular elections. In addition to providing jobs for pollsters, proliferating polls now give everyone the evidence to prove their favorite theory.

However, the swirl of contradictory facts leads me to join the Moody Blues in singing, “Just what the truth is, I can’t say anymore.”

Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982, including Sen. John Kerry (D-Mass.) in 2004.