Sunday, September 18, 2005
New Jersey GOP's big lie on electoral fraud
This is a story that MYDD.com flagged, but a Brookings scholar visiting at George Mason gets credit for busting.
Basically, the GOP counted how many people with the same first name, last name and birthday in New Jersey voted in more than one county. They found 4,400 voters with the same name, born on the same day, and voting in different counties. "FRAUD" screamed the GOP. The NYT printed what the GOP said without the intervention of brain activity.
(continues in comments)
Comments:
<< Home
(continued from post)
Well, Dr. McDonald says he didn't have the New Jersey database to hand. But he had Ohio's. And when he ran the match, he found that many of the matching names had the birthdate 1/1/1800.
If any voters with that birthdate were voting, it would be a scandal.
But, as anyone who knows anything about databases could have predicted, that's what some counties do when they don't have the birthdate: fill in a default. The GOP never bothered to actually call or talk to any of the voters.
The GOP also finds people with the same name and birthdate voting in different states and deceased voters. McDonald addresses these points as well. He points out that there will be many people whose names and birthdays match purely by chance and that there's no evidence in the study that the GOP actually verified that the "deceased voters" hadn't voted by absentee and then expired.
Personally, I doubt they even checked whether the people who voted were actually the same as the people who expired. If John Brown, Sr. dies while living at home with John Brown, Jr., and John Brown, Jr. then votes, it's very likely one can't tell who is who from the records.
But I may be too trusting. For all I know, the GOP just made up the numbers. Or maybe all the cheaters are GOPers. The reader of the NYT article would never know.
For what it's worth, I think the NYT author signaled to an alert reader why this might be bogus. What he did not do is bedrock journalism: wear out a little shoe leather or make a call to connect a database exercise with the real world.
Post a Comment
Well, Dr. McDonald says he didn't have the New Jersey database to hand. But he had Ohio's. And when he ran the match, he found that many of the matching names had the birthdate 1/1/1800.
If any voters with that birthdate were voting, it would be a scandal.
But, as anyone who knows anything about databases could have predicted, that's what some counties do when they don't have the birthdate: fill in a default. The GOP never bothered to actually call or talk to any of the voters.
The GOP also finds people with the same name and birthdate voting in different states and deceased voters. McDonald addresses these points as well. He points out that there will be many people whose names and birthdays match purely by chance and that there's no evidence in the study that the GOP actually verified that the "deceased voters" hadn't voted by absentee and then expired.
Personally, I doubt they even checked whether the people who voted were actually the same as the people who expired. If John Brown, Sr. dies while living at home with John Brown, Jr., and John Brown, Jr. then votes, it's very likely one can't tell who is who from the records.
But I may be too trusting. For all I know, the GOP just made up the numbers. Or maybe all the cheaters are GOPers. The reader of the NYT article would never know.
For what it's worth, I think the NYT author signaled to an alert reader why this might be bogus. What he did not do is bedrock journalism: wear out a little shoe leather or make a call to connect a database exercise with the real world.
<< Home
More blogs about politics.
