The latest polls show President Obama widening his lead over Mitt Romney in Ohio and Florida, crucial swing states, but the Romney campaign rejects the idea that the tide is turning in favor of the Democrats. Republican skeptics say the polls are wrong because the people conducting them are interviewing too many Democrats and too few Republicans, exaggerating Obama's support and underestimating Romney's. Are the major polls skewed against Romney? Here, a brief guide:

How much of a lead do the polls give Obama?
Obama is ahead by six percentage points among registered voters, and by five percentage points among those most likely to show up at the polls, according to the latest Wall Street Journal/NBC News poll. In Ohio, a state that no Republican presidential nominee has won the presidency without, Obama is ahead 53 percent to 43 percent among likely voters, according to a new Quinnipiac University/New York Times/CBS News poll. Real Clear Politics says that the average poll has Obama ahead with 48.7 percent to Romney's 45 percent nationwide, with a five-percentage point lead in Ohio and a three-point edge in Florida. Pro-Romney skeptics say the polls — all of them — are biased in Obama's favor.

What do they think is the problem?
The skeptics say that pollsters are choosing their sample sizes based on the assumption that the 2012 turnout will be similar to 2008's, when record numbers of Democratic-leaning Latino, black, and young voters went to the polls. This year, though, some of those Democrats have grown disillusioned, and Republicans eager to deny Obama a second term are more motivated than they were four years ago, Romney supporters say. "I don't think [the polls] reflect the composition of what 2012 is going to look like," Romney pollster Neil Newhouse says. Some polls even forecast a turnout even more favorable to Obama than in 2008, GOP consultant Rick Wilson says in the New York Daily News, "which even Obama's most dedicated partisans think is highly unlikely."

Why do Republicans think the pollsters are getting it wrong?
They say they smell a rat. "The Democrats want to convince [anti-Obama voters] falsely that Romney will lose to discourage them from voting," Republican pollsters John McLaughlin tells National Review. "So they lobby the pollsters to weight their surveys to emulate the 2008 Democrat-heavy models." One conservative blogger, Dean Chambers, has even created a website called where he adjusts the data according to assumptions by automated pollster Rasmussen Reports, which tends to favor Republicans, and concludes that Romney has led Obama 52 percent to 44 percent in the average poll since Labor Day.

Are the complaints valid?
According to pollsters, they're ridiculous. Poll takers do weight their samples according to demographic traits, such as race, gender, and age. Party identification, though, is something their polls should measure, not control. The nation is roughly evenly divided between Republicans and Democrats, with a smaller group of independents in between, but there's usually a "party-ID gap" at the polls that reflects the mood of the electorate (Democrats had a seven-point edge, 39 percent to 32 percent, in 2008). If pollsters try to even out the numbers artificially, says Doug Schwartz, director of the Quinnipiac University Polling Institute, "they are substituting their own judgment as to what the electorate is going to look like. It's not scientific."

So are the polls unfairly skewed against Romney, or not?
Look, says Jim Geraghty at National Review, "a split of D+7 or more is excessive." A Democratic turnout advantage of three or four percentage points — halfway between the GOP peak of 2010 and the Democratic one of 2008 — is more fair. If pollsters think Dems can count on another historical high, they should at least "explain why." Pollsters say their critics are underestimating the weight of minorities in the electorate. The fast-growing Latino population, for example, favors Obama over Romney, who drifted to the right on immigration during the GOP primaries. Not only that, says Steve Benen at MSNBC, but the notion that pollsters — "literally, all of them" — would be "deliberately publishing bogus poll results is pretty silly." Silly, but predictable, says Doug Mataconis at Outside the Beltway. "The 'biased polling' argument is just a sub-set of the 'biased media' argument" conservatives have flogged for decades. You can bet that if Team Romney's internal polls were better, they'd be leaking them. They're not, because polling is a science, and the trend is clear. If you deny it... well, "That's not skepticism, that's partisan spin."

Sources: The Hill, MSNBC, National Journal, National Review (2), New York Daily News, The New York Times, Outside the Beltway, The Wall Street Journal

Read more political coverage at The Week's 2012 Election Center.