Location: Tulsa, Oklahoma, Tonga

Thursday, June 02, 2005

Too few Democratic criminals?

Seattle Times' David Postman has been live-blogging the gubernatorial election trial in Washington:

POSTED 3:20 PM Wednesday
Democrats' expert witness says GOP analysis flawed

Christopher Adolph, a University of Washington professor and expert witness for Democrats, is on the stand being questioned by Democratic party attorney David Burman.

Adolph is an assistant professor of political science and on the faculty at the UW's Center for Statistics and the Social Sciences.

He is expanding on his testimony last week about what he says are serious flaws in the Republican statistical analysis of illegal votes.

Republicans want illegal votes by felons and others in any given precinct apportioned between Rossi and Gregoire by the same percentage as the total vote in that precinct.

I almost agree with the Republican argument, except that they are too modest. I can think of many very visible Republican criminals (e.g., G. Gordon Liddy, Chuck Colson, H.R. Haldeman, John Ehrlichman, E. Howard Hunt, Jeb Magruder, James McCord, John Mitchell, Donald Segretti, Robert McFarlane , Oliver L. North, Elliott Abrams, Clair E. George, John M. Poindexter, Alan D. Fiers, Jr., Richard V. Secord, Duane R. Clarridge, Casper Weinberger , a laundry list of (mostly) sex criminals , etc.) --and assume that they are a representative sample. Therefore I conclude that a majority of felon-cast votes in any given precinct should be awarded to Rossi. But, back to the trial:

"This method has no pedigree, no claim to accuracy," Adolph said.

Adolph says the Republican's expert report is unreliable because information appeared to be collected mostly from parts of the state that supported Gregoire.

"This strikes me as clear circumstantial evidence of systematic sampling," he said. That might not have been intentional, but "it does suggest that any inferences from the petitioners data alone will be biased because of their methods."

But he said he used the Republican model to apportion illegal votes alleged by both Democrats and Republicans and says there is "approximately zero chance" that the outcome of the election would change.

See? Rossi couldn't get elected with just a proportionate share of felons. He needed to corner the market on crooks and liars.