Public opinion polls show the horserace between Barack Obama and John McCain to be essentially tied. I think all these polls may be inaccurate, and that Barack Obama is probably still ahead. I’ll explain why.

Scientific polling depends on several factors, chief among them a random sample that accurately reflects the demographic makeup, ideological leanings, and impending ballot box impulses of the electorate as a whole. After all, you are allowing roughly 1,000 respondents (in a good poll) to speak for an entire nation of some 300 million souls. So what’s flawed about the samples in 2008? Perhaps, nothing. But I believe they may be consistently under-representing voters aged 18-29—and, believe me, these young voters are living in Obama-land.

There is no way to know for sure about these samples, because few pollsters are transparent with their methodologies. I don’t blame them for being proprietary—it’s a business, after all—but they shouldn’t blame us for being just a tad skeptical. Why would pollsters make such a mistake? A couple of reasons. The first, and easiest to understand, is that young people are difficult to survey because so many of them lack “land line” telephones—and it’s illegal to auto-dial to cell phones. Secondly, most pollsters—like political professionals and journalists of a certain age—are living in the past. They don’t believe that voters in the 18-29 range will turn out in sufficient numbers. I think they are wrong.

If I’m right, the polling samples are compromised in two ways: For starters, they simply don’t have enough young people in them. In addition, the young people they do survey are not a valid cross-section—because they are polling too few people with cell phones. Are young people with old-fashioned phones more conservative (and, thus, more likely to vote for John McCain) than young people as a whole? I think so. I’m not sure why this is so, but I do know that in surveys taken by Harvard’s Institute of Politics, Hillary Clinton did better than Obama among young people with land-lines—and Obama won among those with cell phones. How did I.O.P. pollster John Della Volpe discover this? He conducts his polls online, which I believe is the most effective way to gauge the opinions of young people.

This spring, Reader’s Digest commissioned Della Volpe’s Cambridge, Massachusetts-based firm, SocialSphere Strategies, to poll young voters for us. John has been doing polls of college-age voters for seven years for the I.O.P., and is considered the best in the field when it comes to surveying young people. This poll turned up a couple of intriguing findings: In a head-to-head contest between McCain and Hillary Clinton, the New York Democrat bested the Arizona Republican by 10 percentage points. This probably represents the generic difference between the two parties among young voters at this point in time. In 2004, John Kerry won the under-30 vote against President Bush by nine points. But Obama led McCain in our poll by a whopping 23 points! (I wrote about this in our June issue, which you can find here.) With the election down to its final two months, that number is essentially unchanged; at the Democratic and Republican Conventions, the I.O.P. released its most recent poll, and Obama maintained that same commanding lead over McCain among young voters.

I can hear some of you respond reflexively, “Yeah, sure, but young people don’t vote.” You’re not alone in this view. Prominent Democrat James Carville made that very point recently. So did prominent conservative Jonah Goldberg. Here’s what they may be missing: First of all, the sheer size of the 18-29 cohort makes it a very significant voting bloc. This generation, called “millennials,” are 47 million strong, meaning that even if young voters go to the polls in lower proportions than older voters, it is quite possible that more votes will be cast in 2008 by Americans under 30 years of age than by those over 65. Also, the conventional wisdom about low turnout among young voters may simply be a dated perception. In 2004, under-30 voting participation ticked up to its highest level since the 26^th Amendment gave 18-year-olds the right to vote in 1972. And make no mistake, the Senate is Democratic today because in 2006 voters in college towns in Virginia and Montana went heavily for Senators James Webb and Jon Tester, respectively.

To be fair, as Jonah Goldberg pointed out in National Review recently, every four years various wishful thinkers claim that this is the year the youth vote will be determinative. That could happen in 2008, he conceded, adding, “but I bet it won’t.” Clintonite James Carville went further, scoffing that there was a word in politics to describe candidates who rely on the youth vote—and that word is “loser.”

John Della Volpe begs to differ. “In 2006, the word you could apply to those candidates who successfully appealed to young voters was, ‘Senator,’” Della Volpe told me at the Republican convention. He added: “And the words you might have to use when addressing the national candidate who gets the young voters in 2008 is… ‘Mr. President.’”