Saturday, September 30, 2006

Fired Pollster Revisits 2000 Election

By Lauren R. Dorgan
The Concord Monitor

Thursday 28 September 2006

Former UNH professor and pollster David Moore contends that his new book about the 2000 presidential election - titled "How to Steal an Election" - is not partisan. Moore said he knows it's a "hard sell," but he argues the book simply explains how George W. Bush took the presidency that was rightfully won by Al Gore.

The question of partisanship isn't academic: Moore was fired from his job as a senior editor at the Gallup Poll after he told his bosses about the book last spring. Gallup General Counsel Steve O'Brien said yesterday that writing the book was a "colossally stupid" thing for Moore to do given the polling firm's nonpartisan mission. O'Brien scoffed at the idea that any book with such a title could be impartial.

Moore disagrees.

"I think it's probably as objective as you can be about what really happened," said Moore, who founded the UNH Survey Center in the 1970s and led it until he left for Gallup in 1993. "All I do is present evidence about how an election was stolen."

Moore's book tells the story of election night 2000 by focusing on the folks who crunched the numbers for network television. He argues that the television stations calling Florida - and therefore the election - for Bush in the early morning hours after the election may have been a self-fulfilling prophecy.

Although the networks rescinded the call only a couple of hours later as vote counts tightened, Moore contends that the call made Gore look like a sore loser as he fought for a recount in Florida. That impression influenced not only public opinion but also judicial outlook, making it possible for the U.S. Supreme Court to shut down the recount, Moore argues.

The book kicks off with an explosive anecdote. Fox, the first network to call Florida and the election for George Bush, had a decision team headed by Bush's first cousin, John Ellis. For years, Ellis has maintained that he made the call by looking at the numbers and making back-of-the-envelope calculations.

But Moore has a different account, from the statistician who sat next to Ellis that night. Cynthia Talkov didn't see Ellis making any calculations, Moore writes, but she did hear what he said after he got off the phone with George Bush and his brother, Florida Gov. Jeb Bush.

"Jebbie says we got it! Jebbie says we got it!" Ellis shouted to the Fox decision team, according to Moore's book.

Fox called the race for Bush at 2:15. NBC followed, then CBS/CNN's joint team and finally ABC. Two hours later, they all retracted that projection.

Moore, who said he made no attempt to ask Ellis or Jeb Bush what happened over the phone that night, contends that the president's younger brother "persuaded" his cousin to call the election. It would be naïve to believe Jeb Bush was simply passing Ellis new numbers, he said.

"Why would Jeb Bush care whether Fox called it at 2:15 in the morning or at 8:15 in the morning? Why would he care?" Moore said in an interview. "Why would you care if you know you're going to win, if you absolutely know you're going to win, does it matter whether Fox calls it? Of course not."

Jeb Bush's effort fits into a "broader and sustained effort" to ensure the election for his brother, using some less-than-legal tactics, Moore writes. To make the point, Moore cites a British newspaper's reports on a sweeping and imprecise effort to purge felons from Florida's voter rolls. Democrats were disproportionately cut from the rolls, Moore argues.

Moore further argues that the other networks, blinded by competition, leapt to call the race after Fox. If Ellis hadn't made the first call, he contends, the others would have been more cautious - and would have had time to notice some serious flaws in the data.

But the other networks say they made the decision based on their own reads of the data at 2:15, not based on Fox's call. Moore doesn't believe it. As Moore writes, "none of the network decision teams will admit to being bamboozled by another network."

To prove this, Moore points to his own experience. On election night 2000, Moore observed the team calling the elections for CBS and CNN. Just after 1 a.m., Moore writes, he asked the head of the CBS/CNN team why he hadn't called Florida for Bush.

"Do you want to call the president of the United States on the basis of 30,000 votes?" asked Warren Mitofsky, a veteran of election night operations dating back to 1967.

An hour later, minutes after Fox and NBC had called it, Mitofsky did too. The margin at the time was 51,000 votes, with 179,000 left to be counted and some serious flaws in the data.

Mitofsky, who died this month, always maintained that he was ready to call the election before Fox had. He said he'd already called the control room to let them know the call was probably on its way, according to Moore's book. After the election, Mitofsky and a partner wrote: "We can honestly say that we did not feel any pressure to call the Florida race prematurely."

But Moore disagrees. He said it's because he knew Mitofsky and he talked to people who knew him, who believe he would have held out a little longer and seen the data shift.

Moore warns that the system of calling elections has not been fixed. Far too much blame has been placed on the centralized data-gathering operation, once known as the Voter News Service, and not enough taken by the network decision teams, Moore contends. Among other things, he writes, competition - the rush to be first - still bedevils the need for accuracy in making election calls.

For the record, Bush officially won Florida in 2000 by 537 votes. A media consortium that conducted a manual recount of contested ballots found that the recount Gore sought - limited to a few counties - would have kept Bush the winner, but by a margin of only 225 votes. A full statewide recount conducted under a uniform standard would likely have handed the election to Gore by just about 100 votes, the consortium found.

The book has shipped to some bookstores but is not set for official release until next month. It hasn't been reviewed anywhere yet. Yesterday, it was panned by Gallup's general counsel. O'Brien called the book "embarrassing," and not just because it could hurt the polling operation's image.

"What I saw of it, the theme of the book is so thin," he said. "There's nothing there."

In the interview, Moore characterized his politics as liberal/moderate, though he strenuously argues that it has nothing to do with what's in the book.

Andy Smith, who leads the UNH Survey Center now, said he hasn't had a chance to read Moore's book. It's not unusual to be a Democrat, either in academia or in the world of surveying, Smith said.

But Smith has never heard anyone criticize Moore's work during his decades as a pollster.

"It's impossible not to have political opinions, but it's just a matter of removing those from your day-to-day work," Smith said.

No comments: