If you want to know what people are thinking, THE POLLING REPORT is the place to find it out. I don't know how any association executive can do without THE POLLING REPORT. They do a super job. -- TERRY TOWNSEND, CAE, past president, Texas Hospital Assn.

. . . The future of political ad testing may well lie on the Internet, which up until now has been considered taboo in the realm of campaigns. But in the Virginia gubernatorial race this year, where there was a highly wired electorate, the campaign of Democrat Tim Kaine was able to test potential advertising online using a unique combination of quantitative testing and a new, innovative technology called iModerate. This approach enabled the campaign to not only accurately assess the impact of Republican Jerry Kilgore’s attack ads on the death penalty, and Kaine’s potential responses, with sound quantitative data, but also gave them the ability to garner in-depth, focus group-like responses, which provided valuable insight....

The ability to get high-level qualitative data to elucidate and go beyond the quantitative data gave the Kaine campaign an edge when dealing with its advertising at one of the campaign’s most critical moments. The campaign had always anticipated that Kilgore would attack Kaine over his opposition to the death penalty, and in the fall, that attack came, in the form of two ads. They were emotional, focusing on people whose families had been victimized by murders and who criticized Kaine, saying he would not enforce the death penalty if elected....

Moments like these are true tests for pollsters and consultants because all the hard data in the world can still give pause occasionally when the results are so counter to what is expected. And in heavily pro-death penalty Virginia, these quantitative results were not what was anticipated....

The combined results from the quantitative and qualitative research conducted on the death penalty ads revealed that the campaign’s position was solid. The quantitative findings allowed them to anticipate the impact of the ads on voting intention, while the qualitative findings from the iModerate sessions gave them insight into the deeper impact of the advertising, and why voters were reacting the way they did. This information -- gathered in just two days -- gave the Kaine campaign confidence that the path they had chosen was the right one and provided the guidance needed to strengthen the ads....

___________

. . . When McGreevey announced his resignation in August of 2004, the public did not believe he did so simply because he was outing himself as "a gay American." In fact, in a PublicMind poll only 8% said he was resigning because he was gay. Another 11% said it was because of his extramarital affair. Nearly half said it was because he was one step ahead of prosecutors in one or more other scandals....

In some measure, Codey was responsible for Corzine’s eventual victory. As Codey repaired damage McGreevey had inflicted on the party and the state, he made it seem less necessary to turn to the Republicans as an alternative. He also stepped aside so Corzine could have the nomination unopposed. And in the general campaign season, Codey was still so popular that it seemed too risky for Republicans to attack him as responsible for the state’s problems....

Commentators also pointed to the "independent" vote. Local popular mythology has it that because 55% of New Jersey’s registered voters are unaffiliated with either party that these independents, and not the partisans, truly determine the outcome of Garden State elections. If Corzine appeared comfortably ahead at times, they said it was because the independents inclined to Corzine. If the race appeared to be narrowing, they said it was because independents were inclining to Forrester.

In fact, unaffiliated voters who vote regularly also have a regular party preference. And these turn out in higher percentages than those who are truly swing voters. Those polls that used a narrow definition of party identification (and therefore an expansive definition of "independent" voters) showed independents going for Corzine. Those polls that used a narrow definition of "independent" voters showed independents with no particular candidate preference. Most unaffiliated voters with no party leanings at all would never show up on election day.

No matter. The bias of the media was not liberal or conservative, nor Democrat or Republican. The bias was to have a horse race and a zippy explanation for every twist and turn no matter how statistically insignificant....

As one poll after another released its results in October, media reports were that the race was narrowing, but now it was widening, narrowing again, and widening. It was easy to suspect that some TV ad was having a telling effect. It was tempting to suspect that this policy statement or that embarrassment was making its way into the voters’ minds. It was too dull to say, "many registered voters cannot distinguish significantly between the two candidates and are pretty sure it’s not worth their trouble to bother." It was too rarified to delve into each poll’s likely-voter model.

But the ups and downs of the race were simple. Forrester could not rise beyond the low forties. Corzine fluctuated from the mid-forties to fifty. A chunk of unaffiliated but Democrat-leaning voters went from voting for Corzine as the default candidate, to undecided, to would not bother to vote at all, and back again. In no poll did it appear that Forrester was poised to become a successful residual category for the Corzine candidacy....

Despite this, we saw that social issues were potentially treacherous to Kaine’s candidacy because they could be used to create a liberal caricature. Battleground voters shared Kaine’s concern about the need to raise the quality of public education, but were potentially open to Kilgore’s positions on issues such as gun rights and the death penalty, and while they generally oppose discrimination against gay Americans, they also oppose gay marriage and gay adoption. We knew Kilgore would claim Kaine took the "liberal" position on all of these issues, as well as taxes.

We decided to engage Kilgore on social issues early so we could begin to defuse their power, which the polling showed was overstated in the conventional wisdom. For instance, Kaine’s position on abortion rights was preferred to Kilgore’s and voters were essentially evenly divided on the issue of contracts between gay Virginians. Kaine opposed a bill that banned same-sex contracts, saying it would have made things like car sales between two neighbors who are men illegal just because they are both men. Kilgore supported the bill, saying it was needed to make sure Virginia is not forced to recognize gay civil unions created in other states. (By 42% to 39%, voters preferred Kaine’s position.)

Kaine’s Faith and Values
To this end, we conducted two statewide waves of focus groups in January and February of 2005. Participants were moderate to conservative, white, married, church-going voters who were undecided in the race for governor, but who we assumed were predisposed toward Kilgore. Our goal was to see how voters reacted when Kaine was attacked so we could build our defenses. These voters expressed strong support for the death penalty and generally said they could not imagine voting for a candidate who opposed it. Despite this, when voters learned that Kaine’s position was based on his faith, and that he would nevertheless enforce the law, most of these voters accepted his view and the issue became moot.

We also learned that most voters could not reconcile strong faith and liberalism, so once they learned about Kaine’s faith, they would no longer conclude Kaine was a stereotypical liberal. Despite our best efforts to attack Kaine in the remaining time, once these voters had reached this conclusion, they did not budge from their determination that Kaine was either moderate or conservative.

This information gave us a powerful way to confound what we knew would be Jerry Kilgore’s main strategy: disqualifying Kaine as a candidate by depicting him as an extreme liberal. While we learned voters were going to need Kaine to explain his views on the death penalty rather than deflect the issue, Kaine’s answer resonated with voters. In fact, because voters knew the death penalty had widespread support in Virginia, Kaine earned the respect of many voters for taking a position that was clearly based on principle. As one respondent noted, "He must really mean it. There is no way he will win votes with that position in Virginia." Speaking out on this issue gave Kaine an opportunity to show his integrity and honesty -- two characteristics our polling said were the most important to voters....