A statistical analysis of voting results in three battleground states hit us like a wicked shot of moonshine, the kind of jolt that dazes with potency then disappears as clarity lingers in its wake.
Professor Steven Freeman, a professor of statistical analysis at the University of Pennsylvania with a doctorate from MIT, analyzed election data from Pennsylvania, Ohio and Florida.
In all three states, Bush obtained significantly more votes than the exit polls indicated, with discrepancies ranging from 4.7 percent in Florida to 6.7 percent in Ohio. Freeman wanted to figure out the probability of the statistical anomalies.
He calculated that the odds of such incorrect exit polls in just one state are about 1 million to 1. For it to have happened by chance in all three states? The odds of that, Freeman reports, are 250 million to 1.
“As much as we can say in social science that something is impossible,” Freeman concludes, “it is impossible that the discrepancies between predicted and actual vote counts in the three critical battleground states of the 2004 election could have been due to chance or random error.”
Upon reading that, the hotheads at News Hits were ready to start rounding up street mobs and lead a storming of the White House. But cooler heads prevailed. We went looking for other opinions.
First we found Simon Jackman, who urged us to stay off the street mob idea.
Jackman is president of the Society for Political Methodology, as well as director of graduate studies for the political science and statistics departments at Stanford University. He contends that Freeman’s math is fine, but that his analysis may be based on a faulty premise. Freeman’s numbers only hold value if the exit polls are accurate, and Jackman says there is reason to suspect that the polls are significantly flawed.
“At this stage, we have a puzzling discrepancy but, alas, more than one plausible hypothesis. We need more data and analysis to sort out the possible explanations, and that will take time. But almost every serious scholar and political operative I have spoken to on this issue believes that we’ll discover problems with exit polls, not large and widespread electoral fraud.”
Jackman’s skepticism is born of common sense: You don’t need a statistician to calculate the probability that organized fraud so massive could be pulled off without some evidence of the conspiracy leaking out.
John Allen Paulos, a professor of mathematics at Temple University in Philadelphia, articulated that point well in a posting on his Web site in response to Freeman’s piece. “Fraud is hard to believe for many reasons, one being the widespread nature, extending over different states and regions, of the shift to Bush. The difficulty of concealing a conspiracy grows rapidly with the number of conspirators.”
However, Paulos also sees the possibility of no conspiracy but “many people working independently to subvert the election.” He also considers another possibility: “Tabulating machines and the software they run conceivably could have been dragooned into malevolent service by relatively few operatives. Without paper trails, this would be difficult, but probably not impossible to establish.”
Hard evidence? “Not yet,” says Paulos, who is also a columnist for ABCNews.com and the best-selling author of Innumeracy and other books. But he emphasizes how little it takes to change the outcome of an election. If a half-dozen voters per precinct had changed their choice, Paulos says, Kerry could be president-elect.
The concerns go beyond Florida, Pennsylvania and Ohio.
“Similarly huge differences between the final tallies and the exit poll percentages occurred in 10 of the 11 battleground states, all of them in Bush’s favor,” Paulos says.
Which brings us to our lightning bolt of clarity: “Absent any proof or compelling reasons for the differences between the final tallies and the exit polls,” Paulos writes, “I do not understand why these gross discrepancies are being so widely shrugged off.”
Why does this seem to be a nonissue among so many, particularly in the mainstream press? Is it that people are concerned about being tagged conspiracy nuts if they seek to find out if fraud has just been committed?
This is a legitimate story. Knowledgeable, highly educated, well-respected people have looked at this election and concluded there are some statistically mind-boggling results. One explanation is widespread fraud. The media should be on this like hungry dogs on fresh meat.
So far, they’ve treated it like a bone to bury.Send comments to NewsHits@metrotimes.com