Live long enough, and you get to have diverse experiences. Today I get to defend the head of the CIA for telling the truth.
CIA director Leon Panetta is catching some flak on the Intertubes for telling Voice of America that
even if suspect ballots are discounted, President Hamid Karzai will in all likelihood win re-election.
But if you look at the numbers for 15 seconds, it’s clear that Panetta is simply stating the obvious.
Panetta’s full quote was:
“It’s clear that there was some degree of corruption and fraud involved in the election,” Panetta said. “It’s being viewed now by the commissions involved in counting those votes. I think what appears to be the case is that even after they eliminate some of the votes that resulted because of fraud, that Karzai will still – still looks like the individual who’s going to be able to win that election.”
Here are the numbers, according to the New York Times on September 16:
Current tally:
Karzai 3,093,256 54.6%
Abdullah 1,571,581 27.8%
Valid votes 5,662,758
Ballots being reviewed:
Karzai 1,100,000
Abdullah 300,000
So, if every ballot being reviewed were thrown out, the result would be:
With all reviewed ballots excluded:
Karzai 1,993,256 46.8%
Abdullah 1,271,581 29.8%
Valid 4,262,758
But if half of the reviewed ballots were thrown out (assuming that half of reviewed Karzai ballots and half of reviewed Abdullah ballots are thrown out – that is, throwing out more than 3 times as many Karzai ballots as Abdullah ballots), the result would be
With half of reviewed ballots excluded:
Karzai 2,543,256 51.2%
Abdullah 1,421,581 28.6%
Valid 4,962,758
That is, if half of the reviewed ballots are thrown out, Karzai still has a first round victory. This suggests that Panetta’s statement that, “even after they eliminate some of the votes that resulted because of fraud” it looks like Karzai will win, is probably correct.
Note that, contrary to what some people seem to believe, Abdullah’s people also stole votes, as the Washington Post reported on August 28. So it’s quite reasonable to expect that if many of Karzai’s reviewed votes are excluded, a proportionate number of Abdullah’s reviewed votes will also be excluded.
Assuming again an equal rate of excluding votes among those now attached to either camp, the crossover point is 66%: roughly speaking, to deny Karzai a first round victory, two-thirds of the votes under review have to be thrown out.
The overall result should not be very surprising to anyone who read the Washington Post on August 11. The Post reported:
In a poll released Monday, Karzai led with 45 percent of the vote among decided voters, compared with 25 percent for Abdullah Abdullah, a former foreign minister. The U.S.-government-funded poll by Glevum Associates, conducted July 8-19, had Ghani fourth, with 4 percent of the vote.
So, any of the above results are not far away from what a U.S.-government-poll predicted.