When Mohammad Khatami was elected to his first term as Iran's president in 1997, the election results surprised many Iranian opinion-poll researchers.
Outsider Mahmud Ahmadinejad's two-round presidential election victory in 2005 came as a total surprise to those who relied on Iranian opinion-poll research to follow the race.
Now, just days ahead of the country's June 12 presidential election, some Iranian opinion polls suggest a new trend developing -- with incumbent President Ahmadinejad reportedly falling behind reformists.
At the same time, opinion polls conducted by Iran's state agencies say otherwise.
It's a sobering reminder of the challenges that frequently leave experts guessing up to the very last minute which way Iran's electorate will lean in key elections, including voting this week that has inspired the biggest public rallies since the toppling of the Shah and the establishment of hard-line Islamic rule three decades ago.
One recent survey released by a group that calls itself the Young Journalists Club in Iran suggests that cleric and former parliamentary speaker Mehdi Karrubi will present the main challenge to Ahmadinejad in the first of two possible rounds of voting. But that survey, of voters across Iran, says both Karrubi and the other reformist candidate, Hossein Musavi, are trailing Ahmadinejad.
Meanwhile, an opinion survey of Iran's 10 largest cities by Ayandeh News suggests that Musavi is ahead of Ahmadinejad by a slim margin -- at least in urban areas.
A separate poll by Islamic Republic of Iran Broadcasting -- conducted only among voters in Tehran -- also shows Musavi leading Ahmadinejad among voters in the Iranian capital.
But pro-Ahmadinejad lawmakers dismiss reports of growing support for the reformist candidates. They cite their own poll numbers, which suggest Ahmadinejad will be reelected easily.
Mutual Distrust
Independent experts say the wide discrepancies of Iran's opinion-poll research are primarily due to the nature of the regime in Tehran. Critics dismiss Iran's system as undemocratic and totally lacking in transparency in its opinion polls, allowing most research to be manipulated by those in power.
Female supporters of President Ahmadinejad in Uromie on June 6Independent experts outside of the country say there are so many contradictions within Iranian surveys that researchers should be extremely careful about assessing future election results on the basis of Iranian opinion polls.
Tehran has a history of jailing opinion pollsters whose work shows that ordinary Iranians are at odds with the government on public policy issues.
Tehran-based socialist Hussain Ghazian spent three years in prison after a survey that he helped conduct in 2002 showed that many ordinary Iranians wanted their government to engage in political dialog with the United States.
Ghazian tells RFE/RL that even when Iran's government isn't trying to cover up the results of public opinion research, there are methodological issues that make nationwide opinion polls questionable.
No Man's Land For Pollsters
Ghazian notes that most opinion poll research in Iran is based on telephone interviews. But he says that method doesn't work for measuring the views of rural or many provincial voters because so many of them do not have telephones.
"It is possible by telephone polling in Tehran and in big cities to make predictions about people's voting behavior," Ghazian says. "But the social situation across the whole of the country does not allow for predictions of political behavior that is representative of the entire country. The credibility of theses kind of predictions for the entire country will not be very accurate."
In fact, Ghazian says telephone research on Iran's provincial areas is impossible for nongovernmental organizations, and is, at best, extremely difficult even for state agencies.
A Musavi campaign rally at Haydarniya Stadium in Tehran on June 9."Even in Tehran, where you do have the ability to contact people by telephone, they often will not trust you because they know you have their telephone numbers," Ghazian says. "Their answers to questions -- especially on political opinions -- will be very conservative because they fear they are under surveillance and being monitored. Even for nonpolitical questions, such opinion research is not very precise."
The International Crisis Group also has said there is a large margin of error for public-opinion research conducted in Iran. It notes that as many as half of the people contacted will refuse to answer the questions.
Among those who do respond, there often are strong suspicions about the pollsters -- contributing to a culture in which people say what they think the pollster wants to hear rather than their own views.
Lots Of Room For Error
According to Iran's Interior Ministry, there are 42.6 million Iranians who are eligible to cast ballots in the June 12 presidential election.
Iran's latest general census, conducted in 2007, says that about 38 percent of the country's population lives in rural areas or village communities with no more than 500 residents.
A Karrubi campaign poster in TehranThat census also says about one-quarter of Iranians live in Tehran and the surrounding metropolitan area. That means about one-third of Iran's population lives in towns or urban areas outside of the capital.
Some independent analysts outside of Iran conclude that the candidate who can muster strong support from voters in provincial or rural areas will win this month's presidential election.
While that rural support is difficult to quantify or independently verify, anecdotal evidence does suggest that Ahmadinejad has a strong chance of winning reelection.
But at the end of the day, independent analysts say the only accurate way to name the winner of an Iranian election is to cite the actual election results.
Radio Farda's Mehrdad Mirdamadi and RFE/RL's Mazyar Mokfi contributed to this report
Outsider Mahmud Ahmadinejad's two-round presidential election victory in 2005 came as a total surprise to those who relied on Iranian opinion-poll research to follow the race.
Now, just days ahead of the country's June 12 presidential election, some Iranian opinion polls suggest a new trend developing -- with incumbent President Ahmadinejad reportedly falling behind reformists.
At the same time, opinion polls conducted by Iran's state agencies say otherwise.
It's a sobering reminder of the challenges that frequently leave experts guessing up to the very last minute which way Iran's electorate will lean in key elections, including voting this week that has inspired the biggest public rallies since the toppling of the Shah and the establishment of hard-line Islamic rule three decades ago.
One recent survey released by a group that calls itself the Young Journalists Club in Iran suggests that cleric and former parliamentary speaker Mehdi Karrubi will present the main challenge to Ahmadinejad in the first of two possible rounds of voting. But that survey, of voters across Iran, says both Karrubi and the other reformist candidate, Hossein Musavi, are trailing Ahmadinejad.
Meanwhile, an opinion survey of Iran's 10 largest cities by Ayandeh News suggests that Musavi is ahead of Ahmadinejad by a slim margin -- at least in urban areas.
A separate poll by Islamic Republic of Iran Broadcasting -- conducted only among voters in Tehran -- also shows Musavi leading Ahmadinejad among voters in the Iranian capital.
But pro-Ahmadinejad lawmakers dismiss reports of growing support for the reformist candidates. They cite their own poll numbers, which suggest Ahmadinejad will be reelected easily.
Mutual Distrust
Independent experts say the wide discrepancies of Iran's opinion-poll research are primarily due to the nature of the regime in Tehran. Critics dismiss Iran's system as undemocratic and totally lacking in transparency in its opinion polls, allowing most research to be manipulated by those in power.
Tehran has a history of jailing opinion pollsters whose work shows that ordinary Iranians are at odds with the government on public policy issues.
Tehran-based socialist Hussain Ghazian spent three years in prison after a survey that he helped conduct in 2002 showed that many ordinary Iranians wanted their government to engage in political dialog with the United States.
Ghazian tells RFE/RL that even when Iran's government isn't trying to cover up the results of public opinion research, there are methodological issues that make nationwide opinion polls questionable.
No Man's Land For Pollsters
Ghazian notes that most opinion poll research in Iran is based on telephone interviews. But he says that method doesn't work for measuring the views of rural or many provincial voters because so many of them do not have telephones.
"It is possible by telephone polling in Tehran and in big cities to make predictions about people's voting behavior," Ghazian says. "But the social situation across the whole of the country does not allow for predictions of political behavior that is representative of the entire country. The credibility of theses kind of predictions for the entire country will not be very accurate."
In fact, Ghazian says telephone research on Iran's provincial areas is impossible for nongovernmental organizations, and is, at best, extremely difficult even for state agencies.
The International Crisis Group also has said there is a large margin of error for public-opinion research conducted in Iran. It notes that as many as half of the people contacted will refuse to answer the questions.
Among those who do respond, there often are strong suspicions about the pollsters -- contributing to a culture in which people say what they think the pollster wants to hear rather than their own views.
Lots Of Room For Error
According to Iran's Interior Ministry, there are 42.6 million Iranians who are eligible to cast ballots in the June 12 presidential election.
Iran's latest general census, conducted in 2007, says that about 38 percent of the country's population lives in rural areas or village communities with no more than 500 residents.
Some independent analysts outside of Iran conclude that the candidate who can muster strong support from voters in provincial or rural areas will win this month's presidential election.
While that rural support is difficult to quantify or independently verify, anecdotal evidence does suggest that Ahmadinejad has a strong chance of winning reelection.
But at the end of the day, independent analysts say the only accurate way to name the winner of an Iranian election is to cite the actual election results.
Radio Farda's Mehrdad Mirdamadi and RFE/RL's Mazyar Mokfi contributed to this report
Iran Election Diary
Iran Election Diary
