The newest entry to the lexicon of election disasters comes courtesy of the internationally backed Electoral Complaints Commission. This is the body tasked with sifting through thousands of fraud allegations in Afghanistan’s troubled Aug. 20 presidential vote.
In an otherwise bewilderingly complicated set of documents, the “coefficient of fraud” evokes some new form of geek-chic, some skulduggery committed at the point of a protractor.
The phrase actually refers to the percentage of votes ruled fraudulent within a sample set.
Rather than investigate each suspicious ballot box, the commission decided to investigate sample sets and apply the findings across all the suspicious boxes. The team used the "coefficient of fraud" to deduct the same percentage of votes from each candidate's totals.
But, as with any sampling process – think opinion polls – there is a margin of error.
Given that the resolution process left President Hamid Karzai just 1.7 percentage points shy of winning outright, the margin of error of the "coefficient of fraud" sample could be crucial. So how big is it? The ECC, overwhelmed with press queries Monday night, did not have the margin of error on hand and offered to make it available Tuesday.