Adam Nagourney of the New York Times makes a good effort this morning to attack the idea that a summer frontrunner is invincible. His major concern, of course, is Hillary Clinton and the Democrats. Thompson, Giuliani, and Romney are so scrunched together according to one metric or another that it is hard to identify any one of the three as the GOP frontrunner.

While I applaud Nagourney's attempt to do some pushback on all this Hillary-Is-Invincible stuff, I think his argument is not as strong as it could be. He writes:

Typically, a candidate is adjudged a front-runner because he -- or she -- leads in the polls, has the most endorsements, is ahead in fund-raising, gets the most media attention, draws the biggest crowds and, well, just comes across as a front-runner.

Mrs. Clinton has been helped considerably by the perception in Democratic circles that she has outpaced her competitors at most of the candidate debates.

Yet Mrs. Clinton may be a good example of why the front-runner designation is so ephemeral. Mr. Obama has arguably outpaced her in fund-raising and crowds. Both Mr. Obama and Mr. Edwards have held their own in winning endorsements.

Mrs. Clinton may have the lead in national polls and polls in New Hampshire. But most polls show a tight three-way race in Iowa, where many Democrats consider Mr. Edwards the, um, front-runner. Anyway, polls in Iowa and New Hampshire in the fall do not tell you very much about what is going to happen in January.

The truth is, there is no evidence that the Democratic primary voters have fallen head-over-heels for Mrs. Clinton. And any event that reminds Democratic voters of the lingering concerns about her could topple her from her perch.

I agree with Nagourney's general argument. However, he does not deploy it as well as he could have. His problem actually comes in the first paragraph. He wrongly puts the polls first in his list of reasons of how a frontrunner is identified. Now, this is a laundry list - and an item's position in it does not necessarily matter. The problem is that, at least with Nagourney's list, the polls are largely caused by the rest of the items on the list. So, while Nagourney is attacking the conclusion that Hillary is inevitable, he allows to go unassailed the false presumption that the summer polls are independent of the media dialogue. This false premise about the polls undergirds all of the arguments about Clinton being inevitable.

This is how I would say that a summer frontrunner is created. We start with the fact that voters right now are paying little attention to the race. Not only that, they do not have very much information about the state of the race. Now, this might sound surprising to you, but the reality is that the ways that most voters acquire political information are quite different from the ways that you acquire it. Right now, their ways are not offering them a lot of information. And, as a consequence, they are not thinking about or paying much attention to the race.

[A case in point. I like to talk politics with people I meet in personal life to get a sense of what others think about things. I had a guy over at my place in Chicago to help me take the apartment down, and when I told him what I do for a living - we started talking politics. He's a Republican who is undecided. He's a smart guy who obviously takes his vote choice very seriously. He told me that he was not so sure about Giuliani because he changes his issue positions so frequently. The fact that a smart, considerate voter such as this one could get his wires crossed between Giuliani and Romney points to something about what the average voter is thinking right now.]

So, voters are not thinking about politics much right now. Out of the blue, they get a call from a pollster. They're asked to indicate a preference that they have not really formed just yet. How do they answer the question? They draw upon the available information that they have on the race - which is culled from, to quote Nagourney, who "has the most endorsements, is ahead in fund-raising, gets the most media attention, draws the biggest crowds and, well, just comes across as a front-runner." In other words, they draw from what little they know of the dialogue among political elites. And what is the elite dialogue at the moment? As Dan Balz notes, elites are asking: Can Hillary Clinton Be Stopped? So, polling respondents select Hillary Clinton.

This points to the methodological flaw of using polling data to analyze the state of the race. Polls are valuable to a point - but they really cannot be taken as independent evidence of the state of the race. This is how the media's echo chamber is created. The media talks up one candidate over another. The polls echo this talking up back to the media. The media believes the polls offer independent evidence that justifies its talking up, and proceeds to talk up the particular candidate all the more.

This is why I refuse to ask the question that Dan Balz asked this week. The manner in which voters make a selection will change - because of all the money that candidates have. That's when the real campaign begins, you know. It begins when the candidates start their advertising blitzes. That has only begun - and so the campaign has only begun. The fact that the media has nothing better to talk about in August than the "campaign" does not mean that there is a campaign to talk about.

The campaign - the real one - could change everything. Barack Obama will have something on the order of $60 million to communicate to primary voters. Clinton, of course, will be equally well-funded. But the point is that, as far as the average voter is concerned, the media dialogue is about to be drowned out by the the actual campaign. Right now, the media and political elites are the ones largely influencing polling numbers. Starting next month, the candidates are going to be the ones influencing those numbers. And so, average voters are going to have an opportunity to hear Clinton and Obama. Both of them will have an opportunity to say their piece, and have their piece heard, prior to Election Day. Accordingly, the way in which the average person's vote choice is informed is quite different than the way in which the average person's selection in a July poll is informed.

And so, we are left with the following question about the Democratic primary. It is, not coincidentally, the one that we began asking back when Obama declared. Clinton offers experience and steady stewardship. Obama offers change. Which will Democratic voters prefer? Despite all of the chatter from the pundit classes - the fact remains that we do not yet have an answer to this question.

Now, this is not to say that we cannot yet evaluate which candidate - Clinton or Obama - is more likely to win the nomination. It is not to say that we cannot yet identify who the frontrunner is. The trick is that we have to approach the question differently than we have been. We cannot just sit and marvel at the results that WMUR found in the Granite State. That's just a roundabout way of looking into the mirror. Instead, we have to arbitrate between (a) the message that Clinton will offer, (b) the message that Obama will offer, (c) our estimate of which message Democratic primary voters will prefer. Unfortunately, the media has made very little progress in this arbitration because they have been so hung up by these poll numbers.