Donald Lambro has an article in the Washington Times that discusses the differences between Romney's national numbers and his numbers in the early states. He writes:

All the polls show that Rudolph W. Giuliani is the clear national front-runner for the 2008 Republican presidential nomination, but Mitt Romney, who lags in fourth place in the same surveys, is leading in the first four party-preference contests of the 2008 campaign.

This is a rarely reported and little-noticed political paradox that is taking shape in the Republican presidential race, and which in a volatile and unpredictable campaign cycle, with a large undecided vote, could end up surprising pollsters and pundits alike.

Is it a paradox? A paradox is "an apparently absurd or self-contradictory statement or proposition, or a strongly counter-intuitive one, which investigation, analysis, or explanation may nevertheless prove to be well-founded or true." I could go along with that. There might be a paradoxical quality to Romney's lead. But the question is exactly what does the "investigation, analysis, or explanation" that resolves the seeming contradiction yield?

Lambro implies that the resolution might come from the fact that Romney has an inside track - that the early victories will "flip" the polls in the later states thus bringing consistency among the polls. I would argue - and indeed I have argued - that the resolution of the paradox comes when we remember that Romney has been the only one on the air in the early states of Iowa and New Hampshire (and his father was governor of Michigan). The numbers in these states might well change when Giuliani and Thompson begin their advertising blitzes - as will the national numbers when it is their time, too.

Also, a technical note. Rasmussen has Thompson up over Giuliani. Relatedly, Reuters/Zogby has Thompson within the margin of error of Guiliani's numbers. So, "all" the polls do not show Giuliani as the "clear" frontrunner.