Study debunks myths about the ‘undecided’ voter

The Colorado Statesman

If Obama campaign officials don’t seem to be panicking while everyone else in the political world sounds the alarm over a disappointing performance in last week’s debate, it could be because they’ve bought into the findings from an in-depth survey of undecided voters released just hours before the candidates took the stage.

In what liberal-leaning Project New America described as an unprecedented glimpse into the minds of swing voters in four western states, the results of a poll commissioned by the consultants appear to upend a set of persistent myths about voters who remain undecided a month before the election.

Describing his approach to governing in a swing state, Gov. John Hickenlooper decries negative advertising and constant political bickering at a discussion sponsored by the left-leaning Project New America. “What we’re doing now is we’re depressing the entire product category of democracy,” he said.
Photo by Ernest Luning/The Colorado Statesman

Contrary to supposition, said Project New America’s research director David Winkler in a presentation of the survey results in downtown Denver last Wednesday, undecided voters aren’t uninformed, they aren’t planning to make up their minds based on economic policies, and they aren’t likely to break for the challenger.

Project New America research director David Winkler stands in front of a “word cloud” displaying the reasons voters said they were undecided in the presidential race, according to a poll sponsored by the organization that was unveiled on Oct. 3 in Denver.
Photo by Ernest Luning/The Colorado Statesman

“These people are undecided for a reason,” he said. “They’re not going to be swayed by the news of the day or attack ads, any one particular ad. You might call them committed undecideds — they’ve come to this point and they’re really going to take a measured approach.”


A panel featuring seasoned Democratic campaign manager Mike Melanson, advisor to Senate Majority Leader Harry Reid Rebecca Lambe and 2008 Colorado state director for Obama for America Ray Rivera discuss results of a Project New America poll of undecided voters in four western states at a forum on Oct. 3.
Photo by Ernest Luning/The Colorado Statesman

The results were based on a poll conducted by Democratic-leaning Keating Research, a Colorado-based firm, that surveyed more than 4,500 voters in Arizona, Colorado, Nevada and New Mexico, four states Winkler dubbed “America’s new swing region,” totaling as many electoral votes as Florida or Ohio. Unlike most polls, which wind up sampling just dozens of undecided voters — making the margin of error for that group skyrocket in any one individual poll — the PNA survey amassed results from 503 undecided voters, enough, Winkler said, to draw solid conclusions about the group. According to Keating, the poll’s margin of error is plus-or-minus 4.4 percent.

The profile of the four states’ undecided voters — including those who are leaning toward Obama or Republican Mitt Romney but are without firm allegiance — is more likely to be white, women, well educated and identify as both independent and moderate. (The voters in the poll were 81 percent white, 59 percent female, 72 percent self-described moderates, and 83 percent have attended at least some college.)

Although stereotypes depict undecided voters as less informed than those who have made up their minds by this point, Winkler said the reverse is often true. After all, said panelists on a discussion that coincided with the survey’s unveiling, determined single-issue voters don’t need to know very much to decide which candidate to support, but voters who weigh a variety of factors might.

According to the results, the undecided voters keep particularly well informed about the election — 73 percent said they planned to watch the presidential debates and 42 percent said that the debates could be the single most important factor influencing their vote.

“What they’re waiting for?” Winkler asked. “They’re intending to watch the debates and learn about the candidates,” he said, though he added that more than half said they’re going to wait to decide until Election Day or during the week beforehand, lessening the impact of any one particular debate performance.

Although “jobs and economy” were far and away the most important issue for the undecided voters — mirroring the results of every poll — Winkler said that a follow-up question yielded a surprising result. Asked what the undecided voters were looking for in a candidate, the undecided voters answered overwhelmingly with the words “country,” “leadership,” “integrity” and “honesty.” Any one of those factors “trumps the economy, that trumps the issues,” he said.

“Agreement on issue positions is not as important as someone you can trust,” Winkler concluded, noting that roughly four times as many voters said that “someone you can trust” was the key to winning their vote as “agrees with your stand on the issues.” While his lead was small, more of those surveyed said they felt they could trust Obama than they could Romney.

Although Romney holds an advantage among the undecided voters when asked which candidate would do a better job improving the economy, it’s not by as wide a margin as Obama holds on the question of supporting women’s access to health care and abortion. (Significantly, the undecided voters appear to believe that it won’t make much difference to the economy which candidate wins, but agree by a wide margin with Obama on the question of women’s health and abortion.)

“This audience is incredibly supportive of women making their own decisions when it comes to health care. More than any poll we’ve ever done, this group is not pro-life — if they had been, they would have already made up their mind,” Winkler said.

As for the supposition that undecided voters are likely to break for the challenger — after all, they’ve had four years to decide whether they like the incumbent, so if they were likely to side with Obama, they already would have, goes the reasoning — Winkler said the data doesn’t support that conclusion.

Comparing the two candidates, the undecided voters had a more positive opinion of Obama, viewing the incumbent favorably by 51 percent compared with 45 percent unfavorably, than of Romney, who was viewed favorably by 40 percent and unfavorably by 54 percent. A solid 82 percent thought that Obama was likely to win the election.

“This is why these people are undecided — they’re cross-pressured here between the lack of trust in Mitt Romney, the character issues they have with him, and the sense that the economy’s not great and maybe it’s time to change,” Winkler said.

Ernest@coloradostatesman.com