海角大神

Mitt Romney has a big lead among potential Iowa Republican caucus attendees, according to a released on Wednesday.

When Mr. Romney鈥檚 name is added to the list of potential GOP White House candidates, he sweeps the field, leading with a plurality of 35 percent of the vote. The second-place finisher, former Arkansas Gov. Mike Huckabee, gets about 9 percent, while New Jersey Gov. Chris Christie takes the bronze with a tick over 6 percent.

These results are sure to add to the punditocracy鈥檚 chatter about a notional Romney 2016 run. The Mittster himself for months has said that he has no interest in repeating his losing 2012 effort and that it is time for somebody else, such as his VP candidate, Rep. Paul Ryan of Wisconsin, to take up the Republican flag and try to win the White House.

But earlier this week on Hugh Hewitt鈥檚 radio show, Romney allowed that 鈥渃ircumstances can change." This has produced a burst of speculation about whether the former Massachusetts governor might actually be persuaded to run again, and how he would do if he did, and who his VP choice might be this time, and so forth. Talk is cheap on cable news in late August.

Actually talk is cheap on cable news at all times of the year, but that is another story.

Look, Romney is not going to run for president this time around. He has no incentive to put himself and his family through that experience again and has given no indication that he wants to do anything other than play the role of a party elder statesman.

After the 鈥渃ircumstances can change鈥� comment, he gave the example of what sort of level of change he had in mind: Maybe everybody who runs in the primaries will at some point decide they can鈥檛 do it, get together, and implore Romney to run instead.

Romney puts the chances of this at one in a million, but given how strong-minded and self-centered you have to be to run for president in the first place, we鈥檇 put the odds of all other candidates voluntarily stepping aside at one in a trillion. If it happens, pigs will fly 鈥� and carry passengers.

That said, the new USA Today poll tells us something about politics and 2016. What it tells us is that individual surveys have to be approached with caution. It鈥檚 better if you can average results from different firms.

That鈥檚 because methodologies differ. And the way the Suffolk pollsters constructed their survey looks like it gave Romney an artificial boost, in our nonprofessional opinion.

First, they presented self-identified Republican voters with a list of the usual suspect candidates and asked respondents to pick a first choice. Mr. Huckabee won this heat, with 13 percent.

Then they asked a follow-up: 鈥淭hinking back to your first choice, if I added Mitt Romney to that list, would you still vote for [your first choice], or vote for Romney, or someone else?鈥�

That鈥檚 like shining a light on Romney鈥檚 name. Implicitly, the pollsters are questioning the respondent鈥檚 first choice. The natural reaction of at least some would be, 鈥渙h yeah, he looks good now, maybe he would be better.鈥�

We鈥檙e not in any way questioning the professionalism of the poll designers: They鈥檙e solid and know more about it than we do. But polling is at least partly an art, as well as a science, and it鈥檚 possible their first priority was getting a base-line reading on how the actual GOP candidates are doing in Iowa. Then they threw Romney in at the end just to see what would happen.