Poll analysis: Martinez remains in control

The Albuquerque Journal’s second poll of the governor’s race was released Sunday and found Republican Susana Martinez leading Democrat Diane Denish by 6 points.

Martinez led 47 percent to 41 percent, with 12 percent undecided, the Journal reported. When “leaners” were included, Martinez’s lead grew to seven points – 49 percent to 42 percent – and when those surveyed who voted in both 2006 and 2008 were included, Martinez led 49 percent to 39 percent.

In other words, there’s no way to look at the poll results that’s good for Denish.

So the Denish campaign released its own poll, taken at the same time as the Journal poll, that had Martinez leading by 1 point, 47 percent to 46 percent, with 7 percent undecided. Denish Campaign Chair Ted F. Martinez wrote in a fundraising e-mail that the differing poll numbers mean that “there is a volatile electorate and opinions are shifting on a daily basis.”

That logic appears to be flawed. With the exception of two polls conducted by the same Democratic firm, there has been remarkable consistency in the polls in recent weeks – consistency that has shown Martinez in the uppers 40s or better and Denish stuck in the lower 40s:

• On Friday, a new Rasmussen Reports poll had Martinez leading 51 percent to 41 percent if leaners were included, and 50 percent to 40 percent if they were not. 3 percent favored another candidate and 5 percent were undecided.
• On Sept. 22, Denish released an internal poll that had Martinez leading by 5 points, 49 percent to 44 percent.
• That same day, Martinez released her own internal poll that had her leading Denish by 10 points, 50 percent to 40 percent.
• Even if you go back about a month, the last Journal poll had Martinez leading Denish by 6 points, 45 percent to 39 percent.
• Around the same time, the previous Rasmussen Reports poll had Martinez leading 48 percent to 43 percent.

Methodology

Which brings us back to the new Journal poll. It surveyed 941 registered, likely voters from Sept. 27-30 and has a margin of error of plus or minus 3 percentage points.

The newspaper’s pollster, Brian Sanderoff, was quoted as saying, “when one looks deeper into the numbers, Susana’s lead tends to increase.” He noted that Martinez is close to the “magic number” of 50 percent – the point at which, on Election Day, a candidate wins.

For Denish to gain ground, Sanderoff said, she must “increase the turnout of at-risk Democratic voters significantly.”

That appears to be the turnout model on which the newly released poll from the Denish campaign is based, at least in part. The poll, conducted by Third Eye Strategies, surveyed 600 New Mexicans “who voted in 2008 or registered since then.” The polling memo also states that, “Priority calling was given to voters who voted in a recent gubernatorial election. The sample was stratified by region, gender, and age to ensure a proportional distribution of likely voters.”

It was conducted Sept. 28-30 and has a margin of error of plus or minus 4 percentage points.

The turnout model for the new Denish internal poll appears to be based at least in part on the idea that the Barack Obama wave of 2008 will continue, and that those who have registered to vote since then will also turn out.

Sanderoff’s model appears to be different and his results are more complicated – as evidenced by the fact that he released numbers based on three different scenarios – likely voters, likely voters with leaners included, and those who voted in both 2006 and 2008.

That reflects a new reality: There were so many new voters who turned out in 2008, many pollsters aren’t sure what to do with them. History suggests many won’t turn out for the mid-term election, but that isn’t stopping Democrats from spending big in New Mexico and elsewhere to try to get them to the polls.

The bottom line

Regardless, one fact can’t be ignored: There’s been remarkable consistency among polls from four companies that have surveyed this race in the last few weeks – Sanderoff’s independent Research and Polling Inc. for the Journal, the GOP firm Public Opinion Strategies for the Martinez campaign, the Democratic firm Greenberg Quinlan Rosner Research for the Denish campaign, and the conservative leaning Rasmussen Reports. They all show Martinez in the upper 40s or better and Denish stuck in the lower 40s.

Only Third Eye Strategies stands apart, using a survey model that appears to be optimistic. The company did another poll last month, released about the same time as the Denish and Martinez internal polls, that had the two candidates tied.

At this point, bet on the polls from the other four companies being more accurate, but keep an eye on the Third Eye Strategies polls. The bottom line remains the same as what I’ve said before: The race isn’t over, but Martinez is firmly in control.