Fact check: Statisticians dispute Gore's math on Miami-Dade hand counts

From Brooks Jackson/CNN

VIDEO Listen to the Supreme Court hearing (with CNN Radio commentary on who is speaking) (December 1) Part 1 | Part 2 | Part 3 Part 4 | Part 5 | Part 6 (QuickTime, Real or Windows Media) Watch all up-to-the-minute video of Election 2000     TRANSCRIPTS Dick Cheney on CNN's 'Inside Politics' Al Gore talks with CNN's John King Joe Lieberman on CNN's 'Larry King Live' Sen. John McCain on CNN's 'Larry King Live'     ALSO Palm Beach County ballots reach state capital Despite election uncertainty, Reno confident about future Florida lawmakers recommend naming electors in special session; Democrats blast move     INTERACTIVE Election 2000 races still too close to call Fact check: What Bush and Gore are saying Key players in the dispute Election timeline Previous close elections Not ready for Webster's: What is a pregnant chad? Profiles of U.S. Supreme Court justices     IN-DEPTH The Supreme Court reviews the Florida election case     RELATED SITES Supreme Court of the United States Supreme Court of Florida Judicial Watch Leon County Miami-Dade County Palm Beach County Seminole County

WASHINGTON (CNN) -- What are the odds that a manual recount of ballots in Florida's Miami-Dade County would have put Vice President Al Gore over the top in the state's final presidential tally? Longer than Gore himself figures, according to separate studies conducted by economists at two universities.

Gore's arithmetic was simple: Canvassers counted about one-fifth of the precincts in Miami-Dade County before calling it quits. The partial recount produced a 157-vote gain for Gore that wasn't included in the state's certified total.

Projecting that the remaining precincts would have yielded an equal gain of votes for the Democratic nominee, his legal team argued that county's decision to halt manual recounts cost their candidate a net gain of 757 votes.

But the experts disagree. Bruce Hansen, a statistician at the University of Wisconsin, contends that a 600-vote gain for Gore in the remaining precincts is a highly unlikely scenario.

"Those projections were unreasonable. I don't think any statistician allied with the Gore team or not would have, signed onto that," he said.

Professor Hansen's own study of the Miami-Dade vote examined the polling data precinct by precinct, and concluded that Gore would gain as many as 179 additional votes and as few as 15 if the hand recount was completed.

Hansen says the partial manual recount the county had completed focused on precincts with the strongest Democratic support, while those remaining would have favored Gore less.

Yet another university study estimates a 10-1 chance that Gore would actually lose votes in the remaining Miami-Dade precincts. Peyton Young, an economist at Johns Hopkins University, estimates that Gore could lose as many as 62 votes in the remaining precincts or gain as many 283 votes.

And that's based on very sophisticated methods -- choosing possible variables by chance and projecting each precinct's possible outcomes over and over again.

"About 1,000 times we ran this so-called 'Monte Carlo' simulation, trying to estimate how many votes net for Gore or Bush you would get out of a manual recount," says Young.

If those projections are correct, it means Gore would still need to gain votes in other counties even if Miami-Dade conducts a full hand count. The vice president is asking the courts to add 188 votes gained in the Palm Beach County manual recount, and another 52 votes netted for him after a machine recount in Nassau County. Both figures were excluded from Florida's certified total.

But even if Gore gains all those votes along with a full manual recount in Miami-Dade, Young estimates that odds of overtaking Bush at about 3-1.