Home | Opinion | Columnists | Mark Mellman

Time to play ‘pick a poll’ again

Once again, it’s time to play “pick a poll, any poll,” and tell any story you want.

Barack Obama’s overseas trip was one of the most widely covered events of the campaign. The Project for Excellence in Journalism has been tracking news stories since March and found 51 percent of campaign coverage that week devoted to the tour, making it the second biggest storyline in their archive.

Voters couldn’t miss all the racket. Pew found 62 percent saying they heard a lot about Obama’s trip, making it the second most widely known event of the campaign, coming just behind his securing the Democratic nomination.

But did this event — so intensely covered and so widely known — alter the dynamic of the race? That question was on the lips of pundits and reporters around the globe.

Their answer, the latest version of “pick-a-poll,” depended almost entirely on which particular poll they focused on. Depending on your poll of choice, you could say the trip was a boon, a bust or simply irrelevant.

Gallup aficionados alone had two different sets of polls, with two different results on which to rely.

Those addicted to Gallup’s daily tracking poll saw gains for Obama. What had been a narrow two-point margin on the eve of Obama’s overseas sojourn ballooned into an eight-point lead as he returned from his triumph. The Chicago Tribune’s blog declared, “The European bounce for Barack Obama has lifted the Democratic candidate for President to the greatest advantage he has held over Republican rival John McCain, according to the results today of the newest Gallup tracking poll.”

Readers of USA Today’s blog got a rather different assessment from that paper’s Gallup survey. “The Friday-Sunday poll, mostly conducted as Obama was returning from his much publicized foreign trip … shows McCain now ahead 49-45 percent … In June he was behind among likely voters, 50-44 percent.”

So while one Gallup poll purports to show big gains from the trip, another finds serious losses.

The Economist/YouGov poll “revealed” little change for the Illinois senator, who turned a five-point lead on the eve of his departure into a seven-point margin on his return.

Rasmussen headlined their poll “No Gain” — Obama was two points ahead when he left and one point ahead when he came back.

Big gains, no effect, significant losses — the polls enable you to make any case you choose. All of which underlines two observations.

First, beware of stories told about this or any other election based on poll changes. Look at McCain’s current celebrity attack — working by some polls, faltering by others. So many contradictory tales emerge precisely because there is so much data.

Is there any reason to believe the polls would have converged on a consistent storyline in the wake of events during campaigns past, when many fewer polls allowed fewer conclusions? If there had been 10 polls before and after the Kennedy-Nixon debates, would three have shown gains for Nixon, three advantages for Kennedy and four no effect at all? How would our explanations of earlier elections have changed if more data had permitted different conclusions? If nothing else, the proliferation of polls should increase our skepticism about the tales polls purport to tell.

Second, this episode makes clear that enthralling the media is not the same as moving or even educating the public. Despite the intense focus and widespread awareness, Pew found most voters gained little or no insight into the candidate from his travels — just 15 percent claimed to have learned a lot about Obama’s foreign policy views through all the hoopla. With the public gaze firmly fixed, the media had the opportunity to educate, but once again, din triumphed over substance.

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.