Did Mahmoud Ahmadinejad Steal the 2009 Iran Election?

Complaint: Voter Turnout and Ahmadinejad’s Percentages Showed Suspicious Uniformity

Some Mousavi supporters argued that the reduced variation in voter turnout across provinces indicated fraud. This charge was statistical gimmickry, made possible by a 35% surge in voter turnout:

The [Chatham House Preliminary Analysis] claims that the fact that the variation in participation across provinces has dropped is evidence of fraud. Anyone familiar with elementary statistics knows that the standard deviation of any variable limited to 100% from above would drop as its mean increases. (At the limit, when the mean is 100%, the [standard deviation] would be zero!) So, because the participation rate increased by about 35%, it is hardly surprising that the [standard deviation] fell by 23%.

In addition, while the provincial range had narrowed for the reason this writer explains, it nevertheless remained quite wide: from 63% to 99%.

Many other Mousavi supporters added a variation of this “uniform turnout” argument, asserting that Ahmadinejad’s vote totals showed a suspicious uniformity across provinces:

I continue to find these figures unlikely. There is very little variation in Ahmadinejad’s numbers across provinces, except in two cases. In past elections the numbers have been all over the place.

A Time magazine writer was no less perplexed:

Support for Ahmadinejad was strangely consistent across the country, a real change from previous elections, when candidates drew different levels of support in different regions.

This claim was not supported at all by the vote count. Ahmadinejad’s provincial percentages ranged widely in 2009, from a low of 44% to a high of 77% (see note 1) – the same spread as his 40-73% range in 2005. Nor was the 2005-to-2009 swing in Ahmadinejad’s percentages uniform across provinces: it varied from ‑13% to +35%.

One also wonders what figures this Time writer had in mind when he wrote that “Ahmadinejad squeaked into the presidency in a second round of voting [in 2005]…If the results this time are legitimate, it means two-thirds of Iran’s voters have become more conservative over the past four years.” Ahmadinejad had “squeaked into” the winner’s circle with 61.7% of the vote in 2005, nearly 26% more than Hashemi Rafsanjani and less than one point below his percentage in 2009 – hardly support for the suspicious trend this Time writer claims to have spotted.

Another writer claimed to see suspicious uniformity in Ahmadinejad’s performance across economic and ethnic lines: “The 98 percent correlation in Ahmadinejad’s vote across areas of vast economic and ethnic diversity is inconceivable.” The writer cited no authority, and one cannot imagine what might support such an extreme claim. Ahmadinejad won only 34% of the vote in the affluent north Tehran suburb of Shemiran, for example, but 72% in the working-class south Tehran districts of Pakdasht and Islamshahr (see note 1). Eighty percent of the districts won by Mousavi (38 out of 46) were populated predominantly by non-Persian minorities, while Ahmadinejad did best in heavily Persian provinces. Ahmadinejad polled well nearly everywhere, but economic and ethnic variations nevertheless remained.

Another “statistical” allegation was made by several analysts: “How is it that Mr. Ahmadinejad’s margin of victory remained constant throughout the ballot count?” There are several answers, the first being the simplest: it didn’t. When Iran’s official news agency first announced Ahmadinejad’s apparent victory on election night (in response to Mousavi’s premature “victory” announcement – see below) it reported that Ahmadinejad had received 69% of the 5 million votes counted so far – a percentage that gradually dropped to less than 63% as additional batches of approximately 5 million votes each were reported at roughly 90-minute intervals throughout the night. As the number of reported votes increased, the candidates’ overall percentages naturally were affected less and less by each additional batch. Some analysts found it suspicious that batch percentages did not vary substantially from one to the next, but that was hardly surprising since each batch consisted of vote counts from thousands of polling stations. A few commentators alleged that the reported result for each polling station was very close to Ahmadinejad’s 62.6% final percentage, but those allegations were baseless: the Interior Ministry did not report individual polling station results until several days later. When it did, a cursory examination of ballot box reports would show this allegation had no merit. Ahmadinejad’s percentages from different polling stations ranged from 0% to 100%, as did Mousavi’s, in each case with many thousands of different percentages spanning the full range in between.[18]

Complaint: The Result Is Not Plausible Because It Conflicts Sharply With Many Predictions and Post-Election Analyses

Many Western analysts had assumed that the anticipated sharp increase in voter turnout boded well for Mousavi. This assumption reflected several others, including the widespread belief that many voters had sat out the 2005 run-off election to express their dissatisfaction with both candidates,Ahmadinejad and Hashemi Rafsanjani. That assumption, in turn, was based on a belief held even longer by many analysts: the high percentage vote for Mohammad Khatami in 1997 (69%), and his even stronger showing in 2001 (78%), reflected a “liberal inevitability” in Iran, the eventual opening of Iranian society that would occur once another candidate appeared who deserved the support of this vast but dormant voting bloc. Mousavi appeared to be that candidate. Many analysts also assumed that those who had voted for Mehdi Karroubi and other reform candidates in the first-round 2005 election would vote this time for either Karroubi or Mousavi. Finally, many analysts considered it a myth that Ahmadinejad was strongly supported by rural voters. After all, many rural voters had supported Khatami in 1997 and 2001, and Karroubi in 2005.

The short answer to these chagrined analysts is that none of this matters any longer. The only question now is whether Ahmadinejad won the election fairly – not why Iran’s voters failed to behave as predicted. It is not enough to say, as these analysts essentially do: “The election result was so different from what I’d expected that no explanation other than fraud comes to mind. Therefore, the government must prove that fraud did not occur.” The burden of proof is on those who claim fraud, not on those who deny it. Few would insist on enough evidence to make a major dent in Ahmadinejad’s 11 million vote margin – just something beyond disappointment, suspicion, rumor and conjecture. If hundreds or thousands of ballot boxes were stuffed, surely someone can identify at least one. Which polling stations forced voters to use “false pens” with disappearing ink? Where, exactly, were ballot boxes leftunsealed and open? If any of Mousavi’s on-site observers noticed any of this, why did none of them report it?

Nearly all published reports of election-rigging activities have come from unnamed individuals, whose faces one never sees, recounting serious misconduct by unnamed individuals at unidentified places at unspecified times. Even when allegations are made by defectors who have burned their bridges behind them, they have not identified the wrongdoers or offered other evidence, and sometimes stop claiming fraud altogether. Many reports are so detailed that one can scarcely imagine they could have been fabricated, but the vivid details invariably fail to include any information that would permit the story to be verified.

Why the 2009 election did not conform to analysts’ expectations nevertheless deserves a longer answer as well. Several analysts argued that the strong support for Mohammad Khatami in 1997 and 2001 did not necessarily represent the voters’ endorsement of his reformist agenda:

Mohammad Khatami was not swept to office in 1997 on a tide of liberalism or commitment to any ideological stance, but rather because he appeared to be an honest, charismatic anti-establishment figure and one untainted by official corruption. The fact that he was a black-turbaned seyed, a descendant of the Prophet Muhammad, and a disciple of the late father of the Islamic revolution, Ayatollah Khomeini, likely also played well with the religious masses. The personality and style of the candidate himself, and not merely his policy agenda, was the crucial factor in propelling Khatami to his landslide victory.

By contrast, Mousavi had no clerical credentials, nor even a black turban. Fairly or not, both Mousavi and Karroubi also were tainted by charges of corruption. In his debate with Karroubi, Ahmadinejad charged that Karroubi had accepted bribes and suggested that his comparatively lavish life style may have been financed in less than honorable ways. Mousavi was tainted by his association with Hashemi Rafsanjani and his sons, about whom various charges of corruption had been widely circulated. An important reason for Mohammad Khatami’s success in 1997 had been the perceived contrast between him and Rafsanjani, then the outgoing president, who even then was believed by many Iranians to be corrupt. Mousavi allied himself with the very same person from whom Khatami had carefully distinguished himself. While this alliance did not mean that Mousavi himself was corrupt, it greatly boosted Ahmadinejad’s chances of being perceived as the corruption-free candidate. In a poll conducted on the day before the election, when respondents were asked which candidate was “more honest,” Ahmadinejad led Mousavi by 31%.

In addition, more than a few voters may have questioned Mousavi’s passion for the job, since he had largely dropped out of public life 20 years earlier and had devoted most of his time since then to artistic pursuits, becoming a well-regarded abstract painter in the process. While Mousavi supporters often cited his long absence from public life as proof of his above-the-fray political purity, undecided voters may have seen only a diffident man who had barely been coaxed away from his painter’s easel just months earlier and now “mumbles and rushes through his speeches.”

Nor was it safe to presume that voters who had supported reform candidates in the first-round 2005 election would vote for either Karroubi or Mousavi in 2009. Some analysts argued that Karroubi’s success in 2005 was largely attributable to his promise to spread Iran’s oil wealth among the people – a prospect that appealed to many rural voters who may or may not have supported Karroubi’s reformist agenda. With this plank of his platform diminished in 2009 – in no small part because Ahmadinejad had appropriated it in the meantime – Karroubi was predictably less appealing to many rural voters, whose strong religious convictions might well have led them to Ahmadinejad rather than Mousavi.

Ahmadinejad helped poor and rural voters along this path by visiting nearly every district in the country at least once during his first term, and by spreading oil-funded governmental benefits even more far and wide – development projects in rural areas, cash and potatoes to impoverished farmers, low-interest loans to young married couples and small entrepreneurs, increased salaries forgovernment workers, a law providing insurance to three million female rug weavers. This time-honored political practice probably induced many poor and rural voters to express their appreciation for Ahmadinejad on election day, in much larger numbers than most analysts had predicted.

Complaint: Ahmadinejad’s High Percentage Was Not Believable, Especially in Cities and Opponents’ Home Provinces

Numerous commentators wrote that it raises “red flags” or is “simply bizarre,” or “makes no sense” and “seems odd” that so many Iranians could have voted for Ahmadinejad, especially in certain areas. “No one in their right mind can believe” the results, said Grand Ayatollah Hossein Ali Montazeri, one of Iran’s most highly respected clerics (and a harsh critic of the government even before he had been passed over as Supreme Leader after the death of Imam Khomeini in 1989). Some analysts were even more skeptical because Ahmadinejad’s reported percentage had been so high (62.6%). (Many might draw the opposite conclusion from this: vote-riggers who are said to have total control over the number might be safer to pick a lower percentage – not 50.01%, of course, but something in the range of, say, 54-55% would raise far fewer eyebrows than 62.6%.) The best expression of exasperation came from Farideh Farhi, an Iran analyst at the University of Hawaii, who said she “simply, simply cannot believe” this happened. 

Some pre-election polls had appeared to justify the optimism of Mousavi supporters. In the first of eight pre-election voter surveys by University of Tehran pollsters, this one conducted between May 19 and May 21, Ahmadinejad had held an overwhelming 44% lead over Mousavi (63% to 19%). Asurvey conducted by a Western polling organization during the same time frame also indicated a solid (though smaller: 20%) Ahmadinejad lead. Both surveys, however, had been conducted before or just after the list of approved candidates was announced. By June 1, just half-way through the three-week election campaign, Mousavi had already narrowed the gap in the University of Tehran poll to a mere 9% (39% to 30%). But his jubilant supporters failed to notice, or else refused to believe, that the trend then reversed – especially after the June 3 televised debate between the two candidates. By June 11, Ahmadinejad’s margin in the University of Tehran poll had widened to 30% (57% to 27%), slightly higher than his 29% margin later reported for the next day’s election.

Some skeptical analysts nevertheless felt that other pre-election polls had been more reliable. Ms. Farhi, for example, believed the official vote count had been “pulled out of a hat” because it conflicted sharply with “secret Iranian government polls” disclosed six days before the election byNewsweek’s Mazier Bahari. Those secret polls reportedly showed that Mousavi would receive 16-18 million votes to Ahmadinejad’s 6-8 million – a 21 million vote difference from the vote count later reported by the Interior Ministry. Even “a large majority of [Iran’s Revolutionary Guards] also plan to vote for Mousavi,” Bahari had reported – quite a surprise, most analysts agreed, since that large group had been considered strong supporters of Ahmadinejad. Ahmadinejad’s prospects appeared so bleak even to himself, Bahari had added, that he had “gone into a crouch,” had “barred staffers from talking to reporters,” and was “doling out cash to those who attend his speeches, fueling inflation.” Bahari’s anonymously-sourced report of “secret Iranian government polls,” and his comments, soon were cited by many other writers around the world.

The greatest skepticism focused on certain large cities and the home provinces of opposition candidates. Juan Cole found Ahmadinejad’s majority in Tehran province “so unlikely as to raise real questions,” even though Ahmadinejad had been the mayor of Tehran, had won the province by a 10% larger margin in 2005, and had attracted larger crowds than Mousavi to his Tehran campaign rallies. Three days before the election, for example, Mousavi supporters formed a well-publicized “human chain” stretching across Tehran, estimated to include between 18,000 and “at least 100,000” people. Though it was mentioned in far fewer press accounts, a Tehran rally held for Ahmadinejad on the same day drew a crowd estimated at between 180,000 and 1,000,000 people.

And how, Professor Cole and others wondered, did Ahmadinejad manage to do so well (56%) in East Azerbaijan, Mousavi’s home province? One possibility: thanks to some election-year governmental largesse credited largely to Ahmadinejad, students at the provincial university in Tabriz now could obtain a college degree with courses taught entirely in Azeri, their native language – which Ahmadinejad speaks, as he reminded voters during each of his several campaign visits, sometimes by quoting Azeri poetry.

 Ahmadinejad probably had learned the Azeri language during the eight years he had spent as a government official in two Azeri-majority provinces. Another unfortunate coincidence may have diminished Mousavi’s home-town advantage still further: the Supreme Leader (Ali Khamenei) is also an Azeri. And more: “Mousavi is not only from the same town as Khamenei, but according to locals is actually related to the Supreme Leader.” Azeri voters probably had surmised that Khamenei favored the non-Azeri Ahmadinejad over Khamenei’s fellow Azeri townsman and possible kinsman, Mousavi. At least some of those Azeri voters must have wondered whether they should too. In a poll taken three weeks before the election, Ahmadinejad held a 2-to-1 lead over Mousavi among Azeri voters (31% to 16%), though many voters remained undecided. Even with all this in his favor, Ahmadinejad’s percentage in East Azerbaijan was nearly 12% lower than it had been in 2005.

Skepticism ran just as high in the home province of Mehdi Karroubi. He described his dismal showing in Lorestan as “so ridiculous and so unbelievable that one cannot write or talk about it.” Karroubi’s frustration and suspicion were understandable. His nationwide vote total (330,183) was a small fraction of what he had received in 2005 (5,070,114). A possible explanation: Karroubi was a legitimate contender in 2005, a predictably wasted vote in 2009. For this very reason, Mousavi backers had strongly encouraged supporters of Karroubi and Rezai to vote instead for Mousavi (though Rezai voters were more likely to switch to Ahmadinejad). Their poor showings may reflect the success of that effort. Another possibility: Lorestan voters simply preferred Ahmadinejad. He’d won 50% of their votes in 2005, after all, and now he was the incumbent as well: every president of Iran who has run for re-election has won, always in a landslide.