Monday, September 11, 2006

Tennessee Senate: How Important Is a Poll in Rating a Race?

By Stuart Rothenberg

I began this year rating the Tennessee Senate race between Democrat Harold Ford Jr. and Republican Bob Corker as “Clear Advantage” for the GOP, primarily because of the state’s recent political trends, concerns about how Ford’s race and family might affect his prospects, and assumptions about the appeal of a mainstream conservative such as Corker.

Ford is an appealing guy and an excellent campaigner who never seems to run out of energy. His family and race seem less important to me now than they did a year ago, and if any Democrat can win this race, it’s Harold Ford. Do I need to reassess my evaluation of the contest? To steal from William Shakespeare, “To change the rating of the race or not to change the rating of the race — that is the question.”

Late in August, the Ford campaign released a poll conducted by Peter Brodnitz of Benenson Strategy Group. Brodnitz conducted polling for Tim Kaine’s successful gubernatorial race in Virginia, as well as Rhode Island Secretary of State Matt Brown’s less successful bid for the Democratic Senate nomination in the Ocean State.

In other words, Brodnitz is no slouch. He’s a credible pollster with experience and savvy. It’s always a good idea to be suspicious of polling, but these numbers don’t look cooked to me. One Republican operative familiar with the race thought the survey numbers definitely seemed believable.

And the numbers? Ford’s poll showed him leading Corker by a couple of points, 44 percent to 42 percent. Corker’s negatives had risen considerably since Brodnitz’s February poll, while Ford’s favorable ratings had moved higher since then.

On one level, the approaching national Democratic wave, the Tennessee Senate ballot test and many of the poll’s internals suggest that Ford certainly has a chance to win this race. His own ratings are pretty good, and he has more than $5 million in advertising headed for the airwaves.

As I’ve already noted, Ford is an excellent candidate, and he’s running in a year when voters are unhappy with the president and the direction of the nation. They seem ready for something — and someone — different. Ford, an African-American Democrat who aggressively presents himself as a moderate, certainly fills that bill.

But does the poll offer evidence that my initial assumptions and analysis were wrong? Do the poll’s numbers indicate that Ford has a significantly better chance of winning — not of getting close, but of winning — than my initial analysis concluded? Do I need to move the race to “Narrow Advantage” for Corker or even “Tossup”?

Ford leads in the ballot test by 2 points. If the poll is accurate, that’s not much of a margin, but it does mean that he starts out even or a bit ahead rather than behind, unlike his standing in Brodnitz’s June survey, when Ford trailed, 46 percent to 39 percent.

Ford’s favorability ratings have improved dramatically, from 42 percent favorable/31 percent unfavorable in February 2006 to 47 percent favorable/31 percent unfavorable in June and 55 percent favorable/29 percent unfavorable in August.

By contrast, Corker’s ratings have gone from 16 percent favorable/8 percent unfavorable in February to 43 percent favorable/16 percent unfavorable in June and 48 percent favorable/26 percent unfavorable last month.

Democrats see the name identification ratings as evidence of Ford’s growing strength and Corker’s weakness because the Republican’s unfavorable rating has moved up. But it doesn’t seem that simple to me.

Corker’s negatives have grown because he was attacked by his two primary opponents during the GOP primary, while Ford’s favorable ratings have grown and his unfavorables have stayed unchanged because he did not have a primary and instead has been able to run positive ads about himself and his views statewide.

So while the snapshot presented in the Ford poll seems accurate, I’m skeptical about its predictive value.

Brodnitz’s survey shows Corker trailing Ford among moderates by 2-to-1 and getting less than two-thirds of the vote of self-described conservatives. Is that the likely situation in November, after Corker has pummeled Ford in millions of dollars of TV ads? Remember, those ads only need to rally the GOP base, not convert Democrats into Corker voters.

True, Ford will spend millions of his own dollars building himself up and tearing down Corker, and he may well keep the Republican on the defensive. But Corker already has been attacked in the primary, for his performance as mayor of Chattanooga and for his work for unpopular former Gov. Don Sundquist (R).

Remember, Corker is only now starting to paint Ford as too liberal for the state, and the Democrat can expect that some of his decade’s worth of votes on Capitol Hill will become problematic for him this fall.

Republicans have won the past five Senate races in Tennessee, with the Democrats’ best showing in those five outcomes at 44.3 percent of the total vote (by Bob Clement in the 2002 open-seat race). Democrats carried the state in the 1992 and 1996 presidential races, when Tennessean Al Gore was in the vice presidential slot, but the Democratic ticket couldn’t crack 48 percent of the vote either time. And of course Gore lost his home state as the presidential nominee in 2000.

The fact that Ford is getting 44 percent of the vote now — and 90 percent of the black vote — isn’t hard to believe. Democratic candidates seem to be getting their votes early this cycle, and since I all along have expected the Congressman to get 47 percent or 48 percent of the vote, his poll doesn’t shake my initial assessment of the contest. But if Corker’s ads don’t start to move numbers his way by the beginning of October, I’ll certainly take another long look at this race.

This column first appeared in Roll Call on September 7, 2006. Copyright 2006 © Roll Call Inc. All rights reserved. Reprinted with permission.