1 of 2
Fernando Llano, Associated Press
President Barack Obama waves as he arrives at the Convention Center for the second working session of the sixth Summit of the Americas in Cartagena, Colombia, Sunday April 15, 2012.

Lots of people do not like Mitt Romney, or at least think they don't. That much is clear. Romney's favorability in most polls are at historic lows for this point in a campaign. For instance, an ABC/Washington Post survey released early this week showed him underwater: 47 percent unfavorable, 35 percent favorable.

Romney's net negative ratings are unprecedented. Yet, as Chris Cillizza points out at the Washington Post, history does not put too much weight on this number. Walter Mondale, Bob Dole and John Kerry all had strong positives, and all lost, while Bill Clinton had a barely break even rating, and easily won.

The difference in each case was that the losers were all facing an incumbent with a strong economy at his back, while Clinton faced George Bush in 1992 with an economy nearly stagnant.

In short, history suggests that presidential elections are referendums on presidents, not on the challenger.

Nate Silver at the New York Times does a similar analysis and reaches a similar conclusion. Silver focuses on early and late favorability ratings, and notes that early ratings in April tend not to have much predictive force, while late-stage ratings in September and November do.

"Mr. Romney’s mediocre favorability ratings at this early stage of the race are no death sentence. There have been clear reversals in favorability ratings in the recent past once the general election campaign got under way, such as in 1988 and 1992. At least one recent candidate (Mr. Clinton in 1992) won his election with similarly mediocre early favorability ratings. With that said, it would be foolish to suggest that this makes no difference at all. Mr. Romney would prefer to have a positive rating than a negative one. For that matter, Mr. Obama would prefer to have a clearly positive favorability rating than break-even numbers."

A CNN poll released Tuesday comes up with very different numbers for both men. It finds Romney's favorability climbing, inching barely into positive territory, while the president is wildly popular at CNN, 56-42 percent positive.

That CNN poll showed Romney trailed the president by 9 points, right when the Gallup daily tracking poll showed Romney up by 4 or 5 points.

Why are the polls so wildly different in their results? The answer is that they use very different methods to weight their sample, based on demographics, partisan alignment and even race and gender. Pollsters have lately faced fierce criticism from the right and left, with charges they are cooking the numbers. Democrats say that minorities are being undercounted in the Gallup survey, while Republicans charge that the party ID of the samples is skewed toward Democrats.

Sean Trende at Real Clear Politics picked apart these charges and countercharges on Friday, reaching equivocal conclusions. He does note that at this stage in the election, most polls are still focused on registered voters or adult population, both of which favor Democrats. As they near the election, more pollsters will begin applying likely voter filters. But even then, the models used to predict turnout will be disputed.

More interesting than the heave and pull of daily surveys may be the structural differences between the candidates, which are much harder to change and much more telling. At the Wall Street Journal, Dante Chinni offered an interesting analysis on Friday, noting the widely discussed woman problem Romney is facing, but pointing out that it's really more of a marriage problem. Romney leads among married people of either gender, but struggles to win over unmarried voters of both.

"Up to now, analysts have focused on his need to reach 'women' by repositioning himself on some issues or using his wife as a character witness for him. But these numbers seem to indicate he needs to find a way to win over the young, less-wealthy 'not married.' Those voters have been hard targets for him throughout the GOP primary campaign and reaching them might require different, more dramatic shifts in his campaign."

These kinds of demographic survey results have real implications for campaign strategies and the limits of those strategies, Chinni argues. Day to day favorability ratings, not so much.

Eric Schulzke writes on national politics for the Deseret News. He can be contacted at [email protected].