The latest developments out of Iran

In what Nate Silver is calling the worst damage control effort ever, the Guardian Council admitted that the votes collected in 50 cities exceeded the number of eligible voters in those cities, "only" affecting 3 million odd votes.

Even given that the reported turnout was a historic high - over 80% - that's an indication of fairly massive fraud in those cities. While local turnouts, counted by the number of ballots, of more than 100% necessarily imply fraud, it is not necessary for the number of votes to exceed the number of voters in order for fraud to happen - and to do so is a strong indication of the strength of fraud in those cities. If turnout was about average in those cities (and actually a historic high of ~80-85%), then fraud accounted for more than 15-20% of all votes cast in those cities.

And if that figure held in many areas - with or without red flag overturnouts - turnout may not have been at record highs, and we're seeing the sort of degree of massive falsification that could swing an election so dramatically. And if the Guardian Council is admitting that massive fraud happened, I think the case is now quite materially convincing that the sitting president cheated. Not only that, but that the cheating was extensive enough to make a difference.

If I am convinced that fraud was indeed definitely present, and of an order of magnitude large enough to potentially swing nearly any contested election, I doubt that supporters of the opposition are anything but convinced that it did swing the election, and I hope that a peaceful runoff election, rather than violent revolt, is the outcome of this.

Anecdote and Argument

We are sociable creatures who enjoy telling stories. We learn and teach through parables and fables; it should come as no surprise, therefore, that we turn to anecdotes in laying out arguments for and against something.

The problem is, though, all too often we generalize inappropriately. We fail to see the larger picture, because we're interested in the compelling details in front of us that we can grasp directly. When we have a cold winter in Boone, it does not mean that global warming has stopped. The scope of the data needed to talk meaningfully about global warming is much larger than a season in the life of a single town in the mountains.

The same with education, with market policy, with vitamin supplements, and so on. The more grand the topic, the more important it is that we focus on the larger picture. For example, take health care. In the larger view of things, it does not matter if one patient experiences a three hour wait "because of socialized medicine" or another patient faces a jaw-droppingly unexpected million dollar bill "because the private insurers are greedy." What matters is how well the system works and at what cost.

Whether or not abstinence-only education works is a question that cannot be answered by pointing at Bristol Palin; it can only be answered by studies examining the changes in pregnancy rates and STD frequency in its wake (studies do find abstinence-only education lacking, as it so happens). Enjoy your stories, but before you draw your conclusions, how about holding out for science?