Politicians stand in awe of President Bush's ability to extract dollars from donors: $94.5 million in the 1999-2000 cycle, $274.2 million in 2003-04. In both campaigns, he raised expectations to levels no previous candidate had dared, and then beat them.
Now comes another number that seems to shatter the conventional wisdom.
Political scientist Michael Malbin finds that just 30 percent of Mr. Bush's donors for 2000 came back and gave again for 2004.
That's right: Among the donors who put George W. Bush in the White House, seven out of 10 decided not to help him stay there.
More than just counterintuitive, this fact is roughly the inverse of what rules of thumb and past studies would suggest.
Donor attrition was even worse for the Democrats. Only one-quarter of Al Gore's donors from 2000, and 21 percent of Bill Bradley's, gave to any Democratic presidential hopeful in the 2004 primaries.
(All of the fundraising in both parties, of course, was for the period prior to the conventions. All major-party nominees used public financing for the general election campaigns.)
The Democrats' lack of a sustainable donor base at least makes some sense. Neither Gore nor Bradley ran again. But Republicans might have expected to have had a virtual key to the bank accounts of old donors: an incumbent president, immensely popular within the party, rallying the nation in wartime while delivering on tax cuts, judicial nominations and other promises vital to the party's base.
If this seems like a mystery, the man who crunched the numbers would agree. Michael Malbin runs a small think tank called the Campaign Finance Institute, affiliated with George Washington University. He's one of academia's leading experts on campaign finance.
For this project, he compared the 2000-cycle and 2004-cycle lists of donors giving more than $200 to each presidential candidate. He and his grad students checked for misspelled names and other data-entry problems.
Malbin's assessment so far: "I still don't quite understand what I think I am seeing."
Neither do his colleagues in the small world of campaign finance scholars.
"I watched him do it," says Clyde Wilcox, a political scientist at Georgetown University who has studied the motives of political giving. Some of Wilcox's grad students even worked on the project. "He did a really careful job," Wilcox says.
Wilcox has spent years probing the pyramid networks that generate most of the big political money -- networks of fundraisers or "bundlers" like the Bush campaign's Pioneers, who pledged to round up $100,000 each and Rangers, whose target was $250,000 apiece. In Wilcox's earlier studies, donors and fundraisers claimed re-giving rates that were roughly double what Malbin found. Now Wilcox is revising an upcoming study to include questions for donors who gave in 2000 and disappeared in 2004.
A few possible, or partial, explanations:
• First, a small percentage of donors die, get sick, spend their money on other things or otherwise stop giving. Two dozen of the original Pioneers disengaged for a happier reason: They were appointed ambassadors and went overseas, where it's harder to network.
• Second, some of the missing 2000 donors redirected their money to other pro-Bush causes for 2004. Malbin says a significant number gave to the Republican National Committee -- although not enough to explain all of the donor-list decay.
• Third, the earlier studies could simply be wrong. Many of them used the self-reporting of donors and fundraisers, who may remember their intentions better than their acts -- or have other reasons to misstate the facts.
• Fourth, and most significantly, when the donors from 2000 were asked to give again, they might have just said no. Not because they'd become disillusioned with Mr. Bush, but because the magic of his first campaign in 2000 was gone.
• Or, as Republican fundraising consultant Dan Morgan said, they simply decide, "He's in there [the White House]. He doesn't need my money anymore."
That's a problem that seems to have affected the Pioneer program of $100,000 bundlers.
Wilcox's research over the years shows that big contributions are often based on social or business connections, not politics. That was certainly true of the Pioneers.
Jim Francis, the Dallas lawyer who organized the Pioneers in 1999, said they had a "revolutionary fervor factor -- I want to help make a difference." He recruited Pioneers who were pro-Bush, anti-Clinton, pro-Republican and wildly enthusiastic. "That's what made Pioneers 2000 so spectacular," Francis says.
Many of them were from Texas. One former Pioneer says that was a big part of the appeal in what was "a much more motivating time." The fundraiser hopes to maintain political viability by speaking here anonymously.
After the election, the Pioneer program was turned over to the Republican National Committee. The fundraiser says, "It wasn't as special."
It wasn't for the donors either. For the 2000 race, they gave because a Pioneer -- someone they knew and liked -- asked them to. In 2003, all they apparently got was a direct-mail request and a call from a phone bank.
Jim Francis says the Bush re-election organization seemed to shift emphasis, away from Pioneer-style personal solicitations, and toward galas, direct mail and the Internet.
John Green, a political scientist at the University of Akron, says that the shift in emphasis may explain the high rate of donor decay. He says, "The most likely hypothetical right now is that something changed very dramatically in 2004."
He suggests that donors may have less loyalty now, and can drift away from even the most successful politician. The Bush campaign has used barbeques as its favored "donor maintenance" events. Maybe that's not sufficient anymore.
But consider the flip side: The data also indicate that the universe of political givers may be much bigger than anyone thought. That would explain how, even after losing more donors than most candidates ever enlist, Mr. Bush could raise nearly three times more money in 2004 than he did in 2000.