The internet has not yet reshaped politics in any fundamental way, and its failure to do so constitutes a powerful argument for reform of the television, newspaper, and radio media. In his recent book "Politics Moves Online," George Washington University professor Michael Cornfield claims that "the more activists a campaign ensnares in its network-on-the-Net, the more money and volunteer hours it will collect, and the more voters it will be able to reach through mass media and physical contacts." This has come to sound like common sense, even old news. But what sense this claim has is demonstrably false.

At no point during the recent Democratic primaries could you line up the nine or 10 campaigns in order by size of Email list or web traffic and see a better than chance correspondence to number of contributors, or total contributions, or number of mentions in the media -- and almost certainly not to number of volunteer hours worked - although I don't know how to measure that and doubt Cornfield does either. It's not clear what he means by physical contacts. If he means the size of crowds at candidate events, I suspect he is mistaken. If he means volunteers meeting each other through MeetUps (or at least signing up and RSVPing to say they will do so, which is what can actually be measured), he's being tautological. If he means winning votes, he's clearly wrong, even when looking at the one campaign he discusses, namely Howard Dean's.

The Dean campaign's internet presence was phenomenal and may have brought many people into active politics in a lasting way. I hope so. The grain of truth in the rule Cornfield expounds so carelessly is that the Dean campaign had the largest Email list, the most web traffic, the most money, the most volunteer hours, and the most media mentions. But the media mentions preceded the internet growth that the Dean campaign handled so masterfully. And when the media turned against the Dean campaign, the internet organizing couldn't withstand it - the campaign was mortally wounded in a matter of days.

The Kucinich campaign (for which I worked) had one of the largest (possibly second largest) Email lists, often had the second highest web traffic, for a while had the second largest MeetUps, had the highest percentage of small donors, had the second highest number of donors, but never even approached second place in total money or media mentions (not even media stories about Net campaigning), and certainly not in polls or votes.

Cornfield seems to believe that Net organizing somehow generates media coverage. He said at a recent event at the National Press Club that the media paid attention to the amount of money Dean raised on the Internet. But the media covers money no matter how it's raised. And the media's coverage of Dean's internet prowess was simply one type of story the media could do on the candidate leading in the polls and money - a type of story that, like all of the media's favorite types of campaign stories, avoids policy positions.

I do not want to argue that the media created Dean's internet presence. Several campaigns had good media coverage and never developed comparable use of the Net. But it equally clear that the Internet presence did not produce the media. This can be shown by the timing of events and by comparison to other campaigns.

When Kucinich entered the race last Spring, it was not two weeks before the New York Times' Adam Nagourney suggested he had no chance. To make sure that prediction held true, from March to May, 2003, Nagourney mentioned Kucinich 13 times and Dean 111 times. Yet during those months, polls of registered Democrats showed the two candidates running so close that their levels of support were within the margin of error. The three major television networks mentioned Dean 30 times in May and Kucinich not at all, although the April 23 Gallup poll had Dean at 5 percent and Kucinich at 3 percent. From then on, the coverage only got more unbalanced. The polls, the money, and the tremendous Internet activism followed. The media made Dean, and it unmade him again much more quickly - "Yaaaahoooaugh!"

Cornfield, like other commentators, is telling us the media's story about Dean, but the media always claims to be covering the campaigns it creates and reporting on the natural demise of the campaigns it destroys. Cornfield has not looked closely at the matter himself. He hasn't so much as glanced at any of the campaigns other than Dean's for purpose of comparison. Or if he has, he doesn't say so in his book and didn't say so at the Press Club - where he did say that none of the campaigns' websites used discussion forums (as opposed to Email or blogs) to gain ideas from volunteers - although this was the source of most of the Kucinich campaign's organizing ideas.

Cornfield confesses in his book: "My methods and approach have limitations. Except for the survey research, I could not ascertain the representativeness of the evidence and testimony I encountered and collected. In 'hanging out' with members of a community, I was susceptible to the biases attendant on 'going native'." You don't say.

"The Internet's time has, in fact, arrived in politics," Cornfield writes. No, it hasn't. I hope it does. What the Internet has done thus far is demonstrate great power and then shown that power to be as nothing beside the decisive strength of the media. Grasping that lesson would be as important for our democracy as the Internet's direct accomplishments may become. (David Swanson, electronic mail, April 3, 2004)