More than the sum
Today, for the first time, the powerful Guardian Council of the Constitution, responsible for supervising the conduct of the recent disputed presidential election in Iran, acknowledged irregularities in more than 50 constituencies. The actual “irregularity” in these 50 cities is that over 100% of those eligible to vote are recorded as having done so. It seems it is possible to have too much of a good thing, even good citizenship. Claims of vote-rigging by opponents of the declared winner, Mahmoud Ahmadinejad, sparked protests that have been continuing in Iran since the announcement of his victory on June 13, when supporters of the opposition candidate, Mir-Hossein Mousavi, first took to the streets. Just two recent reports and overviews of the situation in Iran are here at The New York Times and here at Prospect magazine.
The question of whether or not the election results were fixed—which has shaded into the question of whether or not the protests were primarily, or continue to be, about vote-rigging—had fascinated many outside observers. It seems that it has turned out to be an academic one, but it’s been interesting to see the various ways professional and amateur statisticians examined it, and how their investigations were viewed by “Kremlinologists” of the regime.
Many outside believed that Ahmadinejad’s perennial popularity in the provinces and the results of pre-election polls suggested that the published results weren’t completely implausible. Some of those also believed that the results were indeed fixed, but only to a limited degree and not to an extent that changed the final result. Bernd Beber and Alexandra Scacco, Ph.D. candidates in political science at Columbia University, looked at the suspicious last digits of the official election counts.
The US polling site, FiveThirtyEight, that did a fine job of analyzing and predicting polling and voting in the most recent US Presidential elections has a good summary and one of its contributors, Nate Silver, gives vent to his newly-justified skepticism:
For all the complex series of statistics that have been run on Iran’s election, it’s the simplest that might prove to be the regime’s downfall. More people “voted” than were eligible to vote—in a lot of places. The interior ministry admits to 50 such instances out of the 300+ jurisdictions in which Iran tallied results. That is widespread, prime facie and admitted-to evidence of fraud, and I don’t see how the Guardian Council expects people to buy the argument that whatever caused the tub to overflow in those 50 cities was not also tainting the results throughout the rest of the country. The Chatham House report we linked to earlier today found that there were more “votes” than voters in two entire provincesThe Interior Ministry will presumably next try to argue that these were irregularities owing to the mere overzealousness of the Iranian people. Perhaps, as happens with some regularity in the United States, people who thought they were eligible to vote but weren’t nevertheless tried to and weren’t screened properly by elections officials. But this explanation doesn’t hold water—voter eligibility is not a tricky matter in Iran. The Statistical Center of Iran reports that, as of the last census, there were some 47.7 million people aged 18 or older in Iran, which is the voting age in that country. By contrast, the widely-cited figure is that there were some 46.2 million eligible voters. Virtually all people aged 18 or older, evidently, are eligible to vote in Iran, which has very few non-citizens (only about 1.6 million according to official estimates).
This leaves only two possibilities: that there was widespread ballot-stuffing or that the results in some or all areas don’t reflect any physical count of the ballots but were fabricated whole hog on a spreadsheet.
[bold emphases mine]
