Was Iran's election rigged? Here's what is known so far.
The Guardian Council has agreed to a partial recount amid alleged irregularities, including an unusually swift announcement of results and a lack of variation in results between cities, rural areas, and ethnic strongholds.
In the biggest presidential election turnout in the history of the Islamic Republic, some 85 percent of Iran's electorate went to the polls last Friday and gave incumbent Mahmoud Ahmadinejad a landslide victory.
Or did they?
Defeated challenger Mir Hossein Mousavi claims that the official result of 62.6 percent for Mr. Ahmadinejad and just 33.7 percent for him was a "dangerous charade," and has called for a new election. His newspaper, Kalameh Sabz, reported that more than 10 million votes were missing personal identification numbers that made the votes untraceable. He also says some polling stations closed prematurely, preventing some voters from casting ballots.
Many others also suspect the legitimacy of the vote, for a number of reasons:
Results from 39.2 million handwritten ballots came much more swiftly than in previous votes, emerging within hours. Detailed election data typically released has not been made public.
Iran's Supreme Leader sanctioned Ahmadinejad's victory after a day, instead of the customary three.
Ahmadinejad made a surprisingly strong showing in wealthier cities, where he is known to have less support, and in the ethnic strongholds of his rivals. Results from cities and rural areas normally vary, but this time were remarkably consistent.
Farideh Farhi of the University of Hawaii, whose decades of studying Iran has included poring over data from Iranian elections, says the result was "pulled out of a hat." Here's why.
Monitor: How does this election compare to past votes in Iran?
Ms. Farhi: My personal feeling is that Ahmadinejad could not have gotten anything more than 10 million. And I really do have the data from previous elections, each district, how they voted, each province, to make comparisons with these numbers that the Ministry of Interior have come out.
I am convinced that they just pulled it out of their hats. They certainly didn't pull it out of ballot [boxes] or even stuffed ballots, they just made up numbers and are putting it out. It just doesn't make sense.
I do take the numbers of the Interior Ministry very seriously. I pore over them every election. I did it last time in the parliamentary election, to determine the orientations and what they mean. I always do that.
In this election, I am not even going to spend time on this, because of all the [problems].
Monitor: Weren't there party monitors at the polling stations, to watch the count?
Farhi: There were party monitors, and the boxes were all counted, and there were records made, and the information was relayed to the Interior Ministry on a piecemeal basis.
But at one point, immediately after the polls were closed, a very few people, without the presence of any monitoring mechanism, started giving out these numbers. And that's why I think this was brazen manipulation.
It wasn't that they only wanted Ahmadinejad to win. They also wanted to make a case that we can do anything we want to do. And they were, I argue, very much interested in demoralizing this 20 to 30 percent extra voters that are coming in.
They simply are not interested in these people continuing to be interested in politics in Iran. The want them to become demoralized and cynical, because their participation in the Iranian electoral process is extremely destructive for the [Islamic] system ...
What they have not counted on, of course, is a group of people that they essentially think of, for lack of a better word, Westernized wishy-washy liberals, who never stand for anything, would actually be upset that this election was stolen in such a brazen way.
They assumed: 'Ah, you know, we go into the streets, we yell at them, and a couple of shots and they go home and close their doors.'
They knew that they were a minority, and that's why they tried to pull this off. They thought they could bully people, through violence. And they may ultimately be correct. But it seems they have underestimated, not only the crowds, but Mr. Mousavi.
Other red flags
Analysts expected a closer race, if not a reverse of that result, after a final surge in cities across Iran galvanized a large anti-Ahmadinejad vote.
Secret Iranian government polls reported by Newsweek earlier this month estimated that Mousavi would win 16 to 18 million votes, and Ahmadinejad just 6 to 8 million. Those polls found that even the Revolutionary Guard and Iran's "vast intelligence apparatus seem to have come around to this position: a large majority of them also plan to vote for Mousavi," Newsweek reported.
Earlier polls appeared to indicate a stronger showing for Ahmadinejad, who – though under fire for poor economic performance, a surge of inflation, and unemployment – had made 60 visits to Iran's provinces handing out cash and development projects.
The final "official" figures, however, gave Ahmadinejad 24.5 million votes, and Mousavi 13.2 million. That result was a shock for many Iranians and analysts.
The powerful Guardian Council will now be investigating irregularities. Daily protests, riots, and violence have marred the aftermath, and Iran's supreme religious leader Ayatollah Sayed Ali Khamenei – who had very quickly pronounced Ahmadinejad's victory "divine" – on Tuesday called for national unity.
But could there have been the widespread fraud? And what does the perception of a stolen vote mean for the hundreds of thousands of Mousavi supporters – and tens of thousands of Ahmadinejad loyalists – who have taken to the streets?
Farhi says of the 11 million new Iranian voters, she "simply, simply cannot believe" that Ahmadinejad could have won 8 million of them.
"The history of the Islamic Republic is that they never vote for status quo, they always vote for change," says Farhi. "I know people who entered this electoral process, who never voted in the Islamic Republic, and they came in and voted simply against Ahmadinejad."
• Material from Reuters was used in this report.