Suspicious Patterns in the Iranian Election Results.

In examining the provincial breakdown of the Iranian election results, I noticed some strange things.

First, while Mousavi was leading Ahmadinejad by 2 to 1 in Tehran in a poll last week, Ahmadinejad won Tehran 52% to 46% in the reported election results. In an authoritarian state, one would expect polling to understate the popularity of the opposition.

More seriously, the great enthusiasm and strong turnout in the election was presumed to be from those voting for Mousavi and change, but the official turnout in areas that favored Ahmadinejad most strongly was much higher than in areas where Mousavi did relatively well.

Here is the breakdown in the six provinces with turnouts over 90% (91.6% - 99.4% turnouts):

Ahmad. 68, Mousavi 31 Ahmad. 56, Mousavi 42 Ahmad. 46, Mousavi 34 Ahmad. 68, Mousavi 31 Ahmad. 77, Mousavi 22 Ahmad. 73, Mousavi 26

All but two have Ahmadinejad receiving at least 68% of the vote. The unweighted average of these six high turnout provinces was Ahmadinejad 64.7%, Mousavi 31.0%.

In the next six highest turnout provinces, all six gave Ahmadinejad at least 69% of the vote.

Here are the results in the six provinces with turnouts under 80% (63.4% - 79.2% turnouts):

Ahmad. 53, Mousavi 44 Ahmad. 47, Mousavi 50 Ahmad. 65, Mousavi 27 Ahmad. 46, Mousavi 52 Ahmad. 59, Mousavi 39 Ahmad. 51, Mousavi 47

Note that all six low turnout provinces are recorded at 65% or less for Ahmadinejad and five of the six are recorded as voting for Ahmadinejad by less than 60%. The unweighted average of these six low turnout provinces was Ahmadinejad 53.5%, Mousavi 43.2%.

In the provinces with the next six lowest turnouts, four provinces reported 66% or fewer votes for Ahmadinejad.

I find it highly suspicious that the high turnout provinces tended to go stronger for Ahmadinejad than the low turnout provinces. Of course, it is possible that the electorate did not trust the anonymity of the voting process, which might have led to a lower turnout in opposition strongholds.