George Barna, a Ventura-based national pollster, projected earlier this month on the basis of hundreds of interviews with voters around the country that Barack Obama will win the presidential election in November easily unless his campaign “commits political suicide.”

Barna said Sen. Obama had taken a 50-35 percent lead over Sen. McCain, and that to win McCain would have to “sweep all these undecided votes — and then some.” Barna added that this is a “particularly remote” possibility because most of the undecided among likely voters are already leaning toward Obama.

Barna’s results jive with two other national polls released weeks later — the Newsweek poll, which found Obama had taken a nearly identical 51-36 percent lead, and the Los Angeles Times/Bloomberg poll, which pegged Obama’s lead at 49-36 percent among all voters. The best known of all pollsters, Gallup, has the race at a dead even 44-44 percent, but although Barna respects Gallup’s work, he stresses that his interviewing questions better distinguish likely voters from registered voters.

He points to the fact that Obama voters are much more excited about their candidate: 53 percent of Obama voters believe their man will win, while only 31 percent of McCain voters expect the same of their candidate. Many of those McCain voters, Barna said, will never show up at the polling booth.

In a follow-up interview conducted in his office, Barna — who moved his business to Ventura in the early l990s — added that he saw “no reason” to think that John McCain could win the election, and gave a half-dozen reasons why he thought Obama could win, quite possibly by a landslide.

Barna mostly polls for large religious organizations, from liberal groups such as “One,” which is known for its association with Bono, to conservative groups such as “Focus on the Family,” which is affiliated with the prominent right-wing evangelical James Dobson. But even among voters who attend church weekly, the Democratic candidate is doing unusually well and the Republican unusually poorly, with more than 23 percent of actively practicing Christian voters who voted for George Bush in 2004 preparing to cross the aisle and vote for a Democrat.

It’s not just Christians. According to Barna’s research, McCain has lost more than 20 percent support to Obama among a wide range of groups that voted Republican in 2004, including men (22 percent), residents of the South (22 percent), conservatives (20 percent) and “downscale” adults (54 percent).  Downscale adults are voters who have not gone to college and make less than $20,000 a year.

“The nation feels insecure and uncomfortable,” Barna said. “Voters want a candidate they can trust, someone who is steady, someone who can bring us into the new millennium. McCain doesn’t have that persona. People see him as cranky, irascible. That’s the image that comes through the media filter, and that’s not what people want.”

Barna stresses the image we see of candidates in the media may have little to do with reality.

“I’ve worked for a lot of these guys, and I know that the media portrayal is not necessarily accurate,” he said. “Obama is much more liberal than most Americans realize. But most people do not vote on issues. People vote on how a candidate makes them feel. And Obama makes people feel good right now.”

Can we trust a June poll on a November election?

Skeptics will point out that national polls for presidential elections in June have not demonstrated much statistical skill. In four out of five of the last elections, the leader in the June polls went on to lose in November, and that includes the last election. John Kerry had a narrow lead over George Bush in the polls of June 2004, but lost in November in crucial swing states such as New Mexico, Nevada and Ohio.

The biggest swing came in l988, when Michael Dukakis, who had a substantial lead of more than 8 percent in June over George H. W. Bush, went on to lose by nearly the same 8 percent margin.  

In that election, the Republican campaign manager Lee Atwater targeted Dukakis as a liberal, mocked him as a phony elitist, and attacked him with ads linking him to the release of rapist Willie Horton.

Could the same strategy — portraying the Democratic candidate as a pretentious liberal — work again this year? It seems an obvious possibility, because Obama is already considered a liberal by most voters, and is a professor besides, and so vulnerable to charges of elitism.

President Bush’s former campaign adviser Karl Rove launched just such an attack shortly after the Barna interview. Speaking to a reporter for ABC, the famous politico compared Obama to a wealthy country club member with a drink and a cigarette, making “snide remarks about everyone who passes by.”

But Barna doesn’t think that kind of attack will work on Obama.

“Americans sense that something has gone astray in our political sphere,” he said. “One of the conclusions widely drawn — especially by younger voters, — is that attacking one’s opponent more and more viciously as the campaign progresses, or the farther behind a candidate falls in the polls is a show of self-interest, not national interest. Add to that the existing perception that McCain represents the old, tired politics of Beltway insiders, and you get a public that is not particularly interested in hearing the old man criticize the younger man, the white man question the integrity of the black man, or the career politician challenge the newcomer.”

But Barna firmly rejected that kind of attack on Obama. He’s a mild-mannered man with graying hair, and he doesn’t speak in colorful quotes, but the pollster all but pounds the table on this point.

“People are already concerned that McCain represents the old way of doing business,” he said. “They don’t want the next president to be a mud-slinger. They want him to stay above the fray. They’re worried about the future.”

A day after Rove’s attack on Obama, McCain’s chief political adviser Charlie Black was quoted in an interview with Fortune magazine saying a terrorist attack could be a “big advantage” to McCain.

Although many independent election observers believe such an attack could in fact play to McCain’s strength — polls consistently show that one issue in which the public has more confidence in the Republican is in his ability to handle terrorism — immediately McCain disavowed the comment. A second campaign official repudiated it, and Black himself apologized for it.

Was this a deliberate move, despite the backtracking? Barna thinks so.

“The statement may have been a strategic move, intended to remind people that in a time of insecurity, people should support the candidate who is more likely to keep the nation secure,” he said. “Notice that the man who said it was not released from his position, even though the McCain campaign has released dozens of staff over the course of the campaign.”

Why Obama will win the Christian vote

Barna is confident Obama will win because McCain has lost millions of voters from the coalition of voters that helped elect George Bush in 2000 and 2004. McCain still is projected to win 78 percent of likely evangelical voters, and 75 percent of registered Republicans. But in an era when the Republican Party has been severely damaged by the unpopularity of President Bush, Barna points to “a huge degree of support” that has been lost to the Democrat among other constituencies, including Catholics, Protestants and non-evangelical born-again Christians.

“The Christian community in the U.S. has largely shifted its loyalty to the Democratic nominee in this year’s race,” Barna declared. “Among the non-evangelical born-again adults, 52 percent supported President Bush in 2004. Only 38 percent are now supporting Sen. McCain, while 48 percent side with Sen. Obama. And notional Christians, who supported John Kerry by an 11-point margin in 2004, today support Obama over McCain by 26 percent.”

Barna’s poll defines Christian voters differently than other pollsters, categorizing evangelicals not by their attendance in a particular church, such as the Southern Baptists, but by seven tests of belief, including whether a Christian believes in Satan, whether the Bible is accurate in all it teaches, and whether they believe in a personal responsibility to share the faith with non-Christians. By this stringent criterion, evangelical voters — who are universally agreed to be much more conservative than most voters — only represent about 9 percent of the electorate.

Other pollsters, such as the Pew Research Center, who do classify voters by the church they belong to, see evangelical voters as a much larger group — more than 25 percent of American voters.

Barna is unshaken by the contrast to Pew and other better-known pollsters. Although all pollsters have found a rise in support for the Democratic candidate in 2008, he is confident that his results better explain the division within American Christianity. In his analysis, truly evangelical voters are a relatively small group, smaller even than the “Skeptics,” which he counts at about 16 percent of the population. So the Republican candidate can win his base of party members and Evangelicals and still lose by a wide margin.

How McCain could win

Barna, who betrays not a hint of personal preference among the candidates, doesn’t entirely rule out McCain’s chances. He offers a half-dozen possibilities that could allow McCain to come back.

Obama could commit “a major political blunder.” Pollsters now agree that McCain took a lead in March when Obama was tarred by the radical statements of his former pastor, the Rev. Jeremiah Wright. The Republican Party led by as much as 5 percent points in a consensus of national polls, but since Obama disowned Wright, he has regained a substantial lead — larger than any margin registered by John Kerry.

“People will be watching closely to see who Obama associates himself with,” Barna said.

Other possibilities: Sen. Obama’s voters, expecting a landslide victory, could fail to show up. A terrorist attack or national security issue (such as the Bin Laden interview released just before the 2004 election) could remind voters of McCain’s experience in international affairs. “A massive number of people either not currently registered, or registered voters who are not currently likely to vote in November, could actually turn out to vote and select Sen. McCain by a substantial margin,” he adds, with a hint of skepticism.

But despite Barna’s attempts to be even-handed, his research leads him back toward the prospect of an Obama victory, with a good chance of a landslide. Not only would McCain have to win all the undecided voters to pull even, but those who plan to vote for McCain are less committed. Among McCain voters, 59 percent say they are “absolutely certain” to vote for their candidate; among Obama voters, the figure is 73 percent. Among registered voters who are likely to vote, 48 percent of Democrats are “excited” about the campaign, but only 30 percent of Republicans say the same.

Barna points out that in 2000, George Bush was a much more conservative candidate than most Americans realized, but that didn’t become apparent until long after he was elected. Obama could benefit from the same lack of scrutiny.

“The only people who pay attention to the details of the issues are journalists,” he said. “We live in a sound-bite society. We hear the Federal government budget is $3 trillion and we want to understand that in 15 seconds.”

And Barna points out a large number of young adults are newly registered and have “no track record” in a national election. If they show up and vote for Obama as they did in the primaries in states such as Iowa, Virginia, Wisconsin and Colorado, chances of an overwhelming landslide look good.

Nate Silver, a Chicago-based statistician who applies probability theory to polls by running their published results through a computer thousands of times for The New Republic and his site, FiveThirtyEight, backs Barna’s research. Silver lists outcome possibilities by percentage, from low possibilities, such as chance of an Electoral College tie (.023 percent), to high possibilities, such as the chance of Obama winning all the Kerry states (65 percent).

The statistician sees no scenario of a McCain victory at more than a 13 percent probability.

The chance of an Obama landslide?

32 percent.