Reprinted in part from Pelast's book by workingforchange.com.

'This series is part of the WorkingForChange campaign, in cooperation with Martin Luther King III of the Southern Christian Leadership Conference, to prevent the theft of the presidential election of 2004. There is a link included to sign onto the WorkingForChange/King petition.

[In the opening excerpts from Palast's book we learned that five months before the November 2000 election, Governor Jeb Bush of Florida and his Secretary of State, Katherine Harris, moved to purge 57,7000 people from the voter rolls, supposedly criminals not allowed to vote. Almost every one was innocent of crimes -- though the majority were guilty of being African American. Palast's discovery of the ethnic cleansing of the voter rolls, which gave George Bush the White, was front page news in Britain, where Palast reports for the Guardian papers. But the report remained hidden from Americans.]

From Planning to Execution to Inauguration; What They Knew, and When They Knew It

An editor at one of the biggest newspapers in the United States told me, "The committee has decided not to continue printing stories about the presidential vote. We think it's over. We don't want to look partisan."

I thought, what "committee"? And I picked up that I wasn't supposed to ask.

America had, as Katherine Harris requested, "moved on."

But I hadn't.

It was now February, and here's what we knew so far. The British Observer story told us that Harris's elections office had wrongly ordered over 50,000 voters stripped from the rolls, thousands of them wrongly. From the Nation report we knew that Governor Bush's office had barred the registration of another 40,000 -- Democrats by a wide margin. That was the election.

Maybe Governor Bush had simply misread the court orders, and maybe Harris's office had no idea the purge list was wildly wrong; maybe their computer firm ChoicePoint DBT simply flubbed the algorithms. One man's mistake is another man's inauguration. Tough, but no criminal intent.

A loose clue still nagged me. As always, it was the money. When I looked into state files, I discovered that ChoicePoint's DBT was not the first contractor on the job. In 1998, this first firm, Professional Service Inc., charged $5,700 for the job. A year later, the Florida Department of Elections terminated their contract, then gave the job to DBT for a first-year fee of $2,317,800 -- no bidding! Then I found out that indeed there had been an open bid for the job. However, when the offers were unsealed, DBT's was the costliest -- several thousand percent over competitors. The state ignored the bids and grabbed for DBT, in the end signing a deal for more than DBT's original astronomical bid. Hmm.

When I contacted database industry experts about the fee paid DBT by Florida their eyes popped out -- "Wow!" "Jeez!" "Scandalous!" The charge of twenty-seven cents per record was easily ten times the industry norm.

Something else bothered me: It was the weird glee, the beaming self-congratulations, from the ChoicePoint public relations man over my Salon report that 15 percent of the names on his purge list were wrong (even though the error turned around an election). To ChoicePoint, my story was good news: In effect, they said, I reported their list was "85 percent correct." But was it?

The Killer Stats

The list was 85 percent "accurate," said DBT ChoicePoint's PR man, because they used Social Security numbers. That was convincing -- until I checked the felon scrub lists themselves and almost none of them listed a voter's Social Security number. Floridians, until recently, did not have to provide their Social Security number when registering to vote.

Four days after I ran my first report in England, on November 30, 2000, the Bloomberg business news wire interviewed Marty Fagan of ChoicePoint, one of the PR men who'd spoken to me. Based on the big "success" of its computer purge in Florida, ChoicePoint planned to sell its voter-purge operation to every state in the Union. This could become a billion-dollar business.

Fagan crowed to Bloomberg about the accuracy of Choice-Point's lists. The company, he said, used 1,200 public databases to cross-check "a very accurate picture of an individual," including a history of addresses and financial assets.

That was impressive. And indeed, every database expert told me (including DBT's vice president), if you want 85 percent accuracy or better, you will need at least these three things: Social Security numbers, address history and a check against other databases. But over the ensuing weeks and months I discovered:

ChoicePoint used virtually no Social Security numbers for the Florida felon purge; of its 1,200 databases with which to "check the accuracy of the data," ChoicePoint used exactly none for cross-checking; as to the necessary verification of address history of the 66,000 named "potential felons," ChoicePoint performed this check in exactly zero cases.

There was, then, not a chance in hell that the list was "85 percent correct."

One county, Leon (Tallahassee), carried out the purge as the law required. But with doubts in the minds of their in-house experts, the county did the hard work of checking each name, one by one, to verify independently that the 694 named felons in Tallahassee were, in fact, ineligible voters. They could verify only 34 names -- a 95 percent error rate. That is killer information. In another life, decades ago, I taught "Collection and Use of Economic and Statistical Data" at Indiana University. Here's a quicky statistics lesson:

The statewide list of felons is "homogeneous" as to its accuracy. Leon County provides us with a sample large enough to give us a "confidence interval" of 4.87 at a "confidence level" of 99 percent. Are you following me, class? In other words, we can be 99 percent certain that at least 90.2 percent of the names on the Florida list are not felons -- 52,000 wrongly tagged for removal.

Okay, you want to argue and say not everyone tagged was actually removed. Maybe 52,000 did not have their vote swiped, but 42,000 or 22,000. Al Gore "lost" by 537 votes.

Now I was confident the list was junk -- it had to be, because ChoicePoint did not use the most basic tools of verification. But why didn't they? Is ChoicePoint incompetent, hasn't a clue of the methodology for verifying its output? That's unlikely -- this is the company hired by the FBI for manhunts, and the FBI doesn't pay for 90.2 percent wrong.

And why would ChoicePoint lie about it? Their list was bogus and they had to know it. Did someone want it wrong? Could someone, say, want to swing an election with this poisoned list? That's when I went back to a stack of documents from inside Harris's office -- and to one sheet in particular, marked, "DBT CONFIDENTIAL AND TRADE SECRET."

"When the going gets weird," Hunter Thompson advises journalists, "the weird turn pro." In London, I showed this "CONFIDENTIAL" sheet to the ultimate pro, Meirion Jones, producer with BBC Television's Newsnight. He said, "How soon can you get on a plane to Florida?"

Mr. Roberts Does a Runner

Our BBC Newsnight broadcast began with a country-and-western twang off the rental car radio:

After hundreds of lies . . . fake alibis . . ."

Newsnight's camera followed me up to the eighteenth fioor of the Florida Capitol Building in Tallahassee for my meeting with Clayton Roberts, the squat, bull-necked director of Florida's Division of Elections.

Roberts, who worked directly under Secretary of State Katherine Harris, had agreed to chat with me on film. We sat on the reception sofa outside his office. His eyes began to shift, then narrowed as he read the heading of the paper on the sofa next to me: "CONFIDENTIAL."

He certainly knew what I had when I picked up the paper and asked him if the state had checked whether DBT (the Choice Point company) had verified the accuracy of a single name on the purge list before they paid the company millions.

"No, I didn't ask DBT. . . ," Roberts sputtered, falling over a few half-started sentences -- then ripped off his lapel microphone, jumped up, charged over the camera wires and slammed his office door on me and the camera crew giving chase. We were swiftly escorted out of the building by very polite and very large state troopers. [Click here for photos.]

Before he went into hiding and called the Smokies, Roberts whipped around and pointed an angry finger at the lens, saying, "Please turn off that camera!" Which we did -- BBC rules. But he didn't add, "and turn off the microphone," so our lawyers ruled we could include his parting shot, "You know if y'all want to hang this on me that's fine." I will. Though not him alone. By "this" he meant the evidence in the document, which I was trying to read out to him on the run. [You can watch the film of the Roberts run for yourself at: http://news.bbc.co.uk/olmedia/cta/progs/newsnight/palast.ram.]

What was so terrifying to this Republican honcho? The "CONFIDENTIAL" page [Click here to see the title page], obviously not meant to see the light of day, said that DBT would be paid $2.3 million for their lists and "manual verification using telephone calls and statistical sampling." No wonder Roberts did a runner. He and Harris had testified to the U.S. Civil Rights Commission -- under oath -- that verification of the voter purge list was left completely up to the county elections supervisors, not to the state or the contractor, ChoicePoint DBT.

It was the requirement to verify the accuracy of the purge list that justified ChoicePoint's selection for the job as well as their astonishingly high fee. Good evening, Mr. Smith. Are you the same Mr. John Smith that served hard time in New York in 1991? Expensive though that is to repeat thousands of times, it is necessary when civil rights are at stake. Yet DBT seemed to have found a way to cut the cost of this procedure: not doing it. There is no record of DBT having made extensive verification calls. It is difficult for DBT to squirm out of this one. If they had conducted manual verification as contracted, you'd think they would have noticed that every single record on the Texas felon list was wrong.

I took my camera crew to DBT's Boca Raton, Florida, office complex to confront them about the verification calls, but they barred our entry. On our return to London, we received a call from one of their executives explaining that "manual verification by telephone" did not "require us to actually make telephone calls" to anyone on the list. Oh, I see.

Based on this new evidence, BBC broadcast that the faux felon purge and related voting games cost Al Gore at least 22,000 votes in Florida -- forty times Bush's margin of victory as certified by Harris. Quibble with that estimate, tweak it as you will, we now knew the rightful winner of the election. Or at least the British public knew.

Fair Use Notice: This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available in our efforts to advance understanding of environmental, political, economic, democratic, domestic and international issues, etc. We believe this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. For more information go to: http://www.law.cornell.edu/uscode/17/107.shtml. If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner.