I was one of three California state coordinators of Dean house parties.  Still stinging from first-hand experience of a bad grassroots plan in Iowa and from reports of a very well-executed grassroots program in New Hampshire (though the gains weren't enough to overcome the "scream" speech), I felt the Dems had learned some important lessons and wanted to ensure the Dem nominee heeded them well.  It seemed to me that grassroots house parties were not just a way to raise some money, but were a crucial GOTV organizing tool--more important than the Dems' traditional door-step visits by strangers.

Bai writes in today's NYT Magazine article that a strategy of having neighbors talk to neighbors is much more effective than flying in outsiders to canvass door-to-door.  The Republicans used this strategy and turned out more people.  Of course, the Dems did turn out huge numbers and the door-to-door approach does also work.  But, we could have done both and I think we would have turned out many more people.  Also, since a house party program is more decentralized, we could done it in every precinct in Ohio (and across the country)-- not just targeted solid-blue precincts-- and turned out more Dem votes within purple precincts and even red precincts of battleground states.

Why didn't the Kerry campaign use the Dean NH approach of house parties for GOTV?  A top Kerry staffer (one of five who had been with Kerry from the very start of his primary campaign and who claimed he talked with Kerry almost daily on the phone) told me: "To be blunt, this is a fat-cat top-down campaign. The campaign staff doesn't really get grassroots."  Those were his exact words (I wrote them down because I was startled he would admit this--I haven't told ANYONE this quote because I didn't want it to get into GOP hands prior to the election).  He did think a grassroots strategy was crucial, but he may have been among the very few Kerry staffers there at the time to think that way; he and one other staffer were pushing to get me hired and create a real grassroots strategy.  He called me daily with updates.  On the fourth day, he apologized that Mary Beth Cahill was concerned I could be a "Republican mole."  He told her I had been a volunteer with the Dean campaign and that he trusted me based on our phone conversations, but that didn't prove anything to her.  She couldn't imagine hiring someone who lived in California that she'd never met.  Instead, she hired a former Emily's List staffer with experience sending direct mail to big donors, whom Mary Beth had worked with previously.  

Of course, I can't say that hiring me would have made a difference in the 11/2 results, nor that there weren't others who would have been better experts on effective grassroots strategies who might have had more experience.  There are so many factors that go into something as enormous as turning out votes across the country.  But, the way the campaign approached that one hiring decision led me to fear how the Kerry campaign was handling its grassroots strategy.  Of course, the Kerry campaign became a small donor campaign when money poured in through the internet and they discovered how easy and effective that was.  But, it was MoveOn, Dean folks, and others who did that, not any particular effort on the Kerry campaign's part.  Once the Kerry folks discovered this because it was in their face, they started sending out regular fundraising emails.  As Kos pointed out in a diary sometime shortly after 11/2, the Kerry campaign's emails continued to be primarily about fundraising even in the last week--still not "getting the grassroots."

The Kerry campaign did eventually hire a key MoveOn staffer to teach them more about using the internet to inspire the grassroots, but even that was a bit of a late hire.

Back in February when they hired a direct mail big-donor person to coordinate their grassroots strategy, I realized that the 527 organizations like ACT would be conducting the grassroots GOTV campaign not the Kerry campaign.  It worried me that the Bush campaign had one strategic headquarters and the Dems would be relying on various groups who couldn't legally coordinate.  (I wrote a dKos diary on this, but the archive search feature isn't working, so I can't find it -- help me out if you know how to search for February or March diaries).  Matt Bai's article published in the NYT Magazine today also talks about how the 527s weren't allowed to mention Kerry's plans to improve anything; they could only tell prospective voters what Bush had done wrong.  Bai wrote in the NYT Magazine a comprehensive article describing the GOP Amway-style approach back in April; Bai probably is a leading expert now on the pros and cons of these two approaches.  Relying on 527s to do the grassroots organizing was a huge strategic error, in my opinion.  I know ACT did a fabulous job, and beat it's own goals in terms of numbers.  But, given the facts (Bush's low approval ratings, a bungled war and foreign policy, a poor economy with many lost jobs, frustration over health care costs, etc.), we should have done much better.  Had the Kerry campaign been able to raise the same total amount (that the campaign + the 527s did) and directly hired folks like Steve Rosenthal and Steve Bouchard, I bet Kerry would now be forming a transition team.

Why did we have 527s in addition to the Kerry/DNC campaign rather than just one entity like the GOP did?  Because we hadn't been building our base over the past decade like the GOP had, and we had to rely on large donations from big donors, not just the small donations from a smaller base.  The 527s were legally permitted to accept these bigger donations.  The two parties had a pretty even playing field when you looked at the total amounts of money raised, but not when you looked at strategy and coordination.

So, I know that the Kerry campaign eventually did have house parties and even put a lot of emphasis on it.  But, they got there reluctantly and it was too late.  My proposal to them talked about the math-- a percentage of the people who go to house parties tend to host their own.  The sooner you start holding these, the more people you'll reach over a many-month period.  The Kerry campaign didn't get their house party program going for about four months after I had talked with them extensively about the benefits of a program like this for getting out the vote.  The former direct mail staffer who was hired to coordinate the grassroots program became the boss of another young staffer hired to coordinate house parties.  The new house party staffer took about two months to design a poor program.  Everyone who looked at it said it was greatly inferior to the program the Dean campaign had up and running in October 2003.  Former Dean house party coordinators from many states called to offer help (for free) repeatedly.  We were ignored, but our suggestions were mostly implemented... after about three months.  By August, the Kerry campaign had a pretty good house party program in place for fundraising, but not for GOTV.  This was about five months too late, and no longer timely.  

My proposal to the Kerry campaign in February was to organize the house parties by precincts in all states, including the crucial swing states.  These would be neighborhood house parties, not simply fundraisers of our most like-thinking political friends.  The Dean campaign did this in New Hampshire.  Cesar Chavez did this organizing Latino migrant workers decades ago.  I don't get the impression the Kerry campaign ever did this.  I don't get the impression ACT ever did this.  I could be wrong, though, and please comment below if these kinds of neighborhood gatherings did take place.  Mary Beth Cahill sent an email in October asking people to host debate-watching parties, and then call swing-state voters from the party.  Perhaps that was the message sent to folks in safe blue states like California, while folks inside the swing-states got a different message.  I don't know.  MoveOn also sent me messages about hosing phone-bank parties and traveling to swing states.  This reminded me of what the Dean campaign didn't do well in Iowa.  Outsiders may have the energy and want to help, but it's locals that really turn out the votes.  It's ironic to me that Mary Beth Cahill didn't trust an out-of-stater she'd never met on her campaign staff, but ACT thought voters would trust out-of-staters on their doorsteps.

The GOP Amway-style pyramid relies on neighbors but is so top-down that I don't see it working well for Dems.  Dems like creativity and input into how they do things, in my experience.  I thought neighborhood-based grassroots house parties could have been a key strategy that might have been as effective or even more effective than the Amway-style pyramids but it was a strategy that needed many months to be built.  It would be more decentralized than the GOP top-down strategy, but the Kerry campaign would have given local folks the tools and encouragement to make it happen.  With the Kerry campaign not "getting grassroots", instead 527s came in to fill the gaps but they had restrictions that made them much less effective, and the Kerry campaign realized what they needed to do (if they realized it ever) much too late when it was in their face.

I can't know whether the Dean campaign would have done better overall, but it sure seems like their grassroots strategy would have been better.  Many Dean folks seemed to understand that the New Hampshire grassroots strategy was more effective than the Iowa strategy.   They already had the house party tools and could have easily re-structured it to focus on precinct-level work for the general campaign.  They already had the small-donor fundraising tools.  Clearly, the Dean campaign would have needed to make other major changes in staff and strategy (especially media and message-control strategy) from the primaries to the general campaign.

Here's hoping the Dems do learn from 2004 as we go forward.  A key lesson is to build our base so that one entity can coordinate a presidential campaign.  Not so that it's top-down, but so that locals in crucial states can talk positively about their candidate and influence their neighbors.