A Big Day for Pollsters and Poll Analysts, Too  | American Journalism Review
 AJR  Features
From AJR,   August/September 2012

A Big Day for Pollsters and Poll Analysts, Too   

Their handiwork will come under intense scrutiny as the votes are counted Tuesday night. Mon., November 5, 2012.

By Rem Rieder
Rem Rieder (rrieder@ajr.umd.edu) is AJR's editor and senior vice president.      


Tuesday isn't just a huge day for President Barack Obama and Republican challenger Mitt Romney.

Also on the line are the reputations of the pollsters as well as the journalists, pundits and political players who interpret those polls. And boy have there been a lot of polls, and a lot of flogging of those polls.

Almost everybody agrees that this is a very close election as far as the popular vote is concerned. And that's reflected by virtually all of the national polls.

But the polls in the critically important swing states tell another story. They suggest that, by a narrow margin, Obama has a lead in a majority of them. And that has led to the widespread belief among political savants that Obama is likely to pile up enough electoral votes to win a second term.

Makes sense. And Obama's strong performance in the aftermath of Hurricane Sandy has certainly given him a finishing kick. But there's no shortage of red flags. While perhaps more likely to happen than not, that's hardly a prediction to take to the bank.

The reason: The size of Obama's edge. In many states, it's within the margin of error for many polls. According to Real Clear Politics Monday afternoon, Obama's lead in Virginia was .2 percent, Colorado .6 percent, New Hampshire 2 percent, Iowa 2.4 percent, Nevada 2.8 percent. And in Ohio, said to be the key element in Obama's all-important firewall, it was just 3 percent.

True, these are averages of all polls, so the likelihood of being wrong is far less than if we were talking about a single poll. And in many of those states, Obama has consistently held that narrow lead. But it's close enough that going the other way is not out of the question. What matters is who shows up. And deciding who is a "likely" voter is not a matter of absolute science. If there is a Democratic enthusiasm gap, if some of those likely voters turn out not quite so likely, it could be a different ballgame.

Also, there's the wild card that is Pennsylvania. That's been parked in Obama's basket throughout the campaign, and it hasn't been promoted officially to swing state status. But the Real Clear Politics average has the president now up in the Keystone State by just 3.9 percent, smaller than his margin in swing state mainstay Wisconsin (4.2).

Yes, I know, Pennsylvania has been GOP fool's gold for years. Just ask John McCain. It hasn't gone Republican since 1988. Like a film noir temptress, it has raised the hopes of many GOP contenders only to lure them to their doom.

But the state sure has tightened. If it flips (again, not likely) and a handful of those swing states swing the other way, it will be time for Romney to start measuring for curtains at 1600 Pennsylvania.

There's no doubt that polling and poll analysis are much more sophisticated than in the past. (Speaking of sophisticated poll analysts, the New York Times' Nate Silver says there's an 86.3 percent chance that Obama wins.) There's a very well-supported case that Obama has the path to 270.

But there are enough wildcards that journalists would be wise to toss in a however, comma and an asterisk or two.

###