By Mahrokh Gholamhosseinpour
December 31, 2020
Between October 22 and 28, 2020, an independent poll conducted for IranWire by Stasis pollsters returned some unexpected results. One was that close to 37 percent of respondents said if they do participate in the June 2021 Iranian presidential election, at the moment they would vote for former President Mahmoud Ahmadinejad.
Other figures whose names have recently surfaced in the context of would-be candidacy, including judiciary chief Ebrahim Raeesi, former president Mohammad Khatami, foreign minister Mohammad Javad Zarif and Mohammad Bagher Ghalibaf, the former Tehran mayor and current speaker of the parliament, lagged behind the ex-president by comparison. And only 25 percent of respondents said they approved of President Rouhani’s performance to date.
Arash Ghafouri, the CEO of Stasis Consulting whose training background includes political management and election campaigns, tells IranWire this data was gathered with precision and by following scientific standards. IranWire interviewed him to gain a better insight about the findings.
Considering that this survey was done through phone interviews, how confident can we be that the respondents have trusted the pollster and the answers have not been affected by misgivings?
No survey can claim its findings are one hundred percent correct or that all respondents have expressed their true views. We can only reduce the possibility of error by following scientific procedures, to make sure we receive the least possible number of erroneous responses.
What procedures are you referring to?
For instance, the pollster must have been well-trained enough to know how to start his questions, how to communicate with the other end and how to earn the respondent’s trust. All those who conducted the poll had been drilled over and over again beforehand, tested the questionnaires, fixed its problems and tried to create a good, intimate atmosphere between the interviewer and the interviewee. But we also have two more criteria to confirm our data. At the end of each interview we ask the respondent to answer two further questions: how much they have trusted the pollster and how honestly they have answered the questions.
And if they have not trusted the pollster?
Those who score low on a five-level scale of “zero to completely” — if, for instance, they say that they have not answered honestly or that they have not trusted the pollster — are removed from the sampling. During this particular survey we removed 42 respondents from our sampling because the assessment was that the respondent had not honestly answered the questions or had answered in a “whatever” manner. In this way we tried to include only the most reliable answers.
How were the participants in this survey chosen?
They were chosen randomly based on gender, age, where they reside and according to the data gathered by the 2016 census. They were interviewed on the phone and all participants were 18 years old or older.
Considering Iran’s population and the number of its provinces, is a sample of 1,136 individuals sufficient for a single-topic survey?
Yes, this number is usual and sufficient. The findings of this survey can be extended to all individuals aged over 18 who reside in Iran with a confidence rate of 95 percent and a confidence scale from 2.8 to 3.1 (for the highest and the lowest possible variance). Sampling of such a number is common across the world for various surveys. Even in the United States, despite its having a population of 300 million, surveys are done via a statistical sampling of around one thousand respondents. Of course, the assumption is that you have chosen your samples correctly. We had access to phones across every province which gave us the ability to choose our samples in each province. It was not as if we had many samples in Tehran but none in Alborz.
How did you accurately choose the distribution of samples in each province?
To arrive at more reliable data and to cover the whole country we chose our samples based on the population of each province, meaning that the number of respondents is proportionate to a province’s population. For instance, Tehran has the biggest population so more respondents were chosen from Tehran.
Perhaps some people wonder how Ahmadinejad or Raeesi have become more popular. Any views on that?
In recent years Mr. Ahmadinejad has presented a new image of himself to the public, one that stands against the ruling oligarchy: a kind of champion of the downtrodden and the oppressed. He did this before, during his presidency, by giving out cash subsidies. He has been trying to make himself into a symbol of opposition to unpopular figures. In the survey, for instance, you can see how unpopular the Larijani brothers [former parliament speaker Ali Larijani and Sadegh Larijani, former head of the judiciary] are. In fact, the fight between the Larijanis and Ahmadinejad and Ahmadinejad’s claim that “I am standing against those who want to plunder taxpayer money” is what has worked to his benefit.
What about Raeesi?
Despite every accusation directed at him, Mr. Raeesi received 13 million votes in the last presidential election and since than he has benefited from a good quantity of favorable news by the media. For instance, stating that he is reforming the judiciary, that he is pursuing justice or that he is fighting financial corruption. Regardless of whether these reports are real or not, or whether he is really after the embezzlers or not, the image of Mr. Raeesi, as portrayed by the media associated with the regime, has helped him to a degree.
Nevertheless, notice that this survey has negative results for both Ahmadinejad and Raeesi as well. For example, the number of respondents who said they like Mr. Raeesi is equal to those who said that they did not like him: 27 percent like him and 27 percent do not. Or, looking at different social strata, a lower number of respondents with a university education who read more news about Raeesi or Ahmadinejad favor them.
Can we draw any conclusions from this survey about the rate of participation in the 2021 presidential election?
When it comes to the rate of participation, this survey is only a snapshot of a juncture in time. We have no idea what the situation is going to be next week, next month or in the days before the election. As usual, however, the proportion of people who say they are going to vote will be higher than the people who actually do vote. Right here in the United States, more than 70 percent said that they were definitely going to vote but after the presidential elections it turned out that only 66 percent had voted.
Is there a way to determine the number of people who say they would vote but do not, and distinguish them from those who are really determined to vote?
We try to use a few complementary questions and a series of statistical models to get closer to the real number of likely voters, but in this survey we did not use them because it is too early to say. The closer we get to the elections, the more we will try to use such questions to get closer to the real likely number of voters which, usually, is fewer.
Would you describe these statistical models a little?
For instance, if given the three choices of “I would certainly vote”, “I might vote” and “I will not vote”, somebody picks “I would certainly vote”, we would ask them if they know on what date the election is going to be held. Naturally, somebody who passionately talks about voting but does not know when the election is to be held cannot necessarily be counted as a person who is absolutely going to vote. We will use these tools in the upcoming surveys as we get closer to the election.
Will these surveys continue until the 2021 election?
Our intention is to continue if we can.
According to the results of this survey, 24 percent of people say that voting is “totally effective” in combating the country’s problems while another 24 percent say that it is “not effective at all”. Are there opinions between these two extremes? Does the survey also reflect the opinion of the 24 percent who say they are not going to participate in the election?
Yes, it reflects all categories. But about those “in-betweens”: in a survey, when you have both unqualified and qualified answers, the unqualified answers become more important because those who have given unqualified answers are more likely to insist on their views than those who have answered “up to a point” or “possibly”.
But there is also another point to make about the effectiveness of voting. You might not want to participate in the election but you might believe that it does make a difference and you might vote depending on the prevailing conditions. If your favorite candidate participates you might believe more in the effectiveness of voting, but if you do not support any of them you might question the effectiveness of participating.
Considering the amount of time between now and the election, we cannot provide accurate answers to questions about the rate of participation in the election. We have to get closer to the election to be able to provide more accurate analyses. But we asked this question because we wanted to have a general idea of the situation so we can compare it to what we find in the future.
Did this survey reveal anything that was new to you, personally?
Yes. In 2015 we did a survey about the popularity of Mr. Ahmadinejad and the answers showed a sharp divide between two extremes. We had asked people how, on a scale of zero to 20, how would they grade Ahmadinejad. Those who graded him zero and those who graded him 20 were in the majority. This time, however, we found fewer people who totally loved or hated him. The number of those who, in the past, had a significant problem with him has fallen and the number of those in the middle bracket has increased. In the same period, people like Larijani have now become very unpopular and even Mr. Rouhani’s approval rating has fallen to just 25 percent.
In 2016, at the height of attention on the nuclear treaty, politicians such as Zarif, Khatami and Rouhani were highly popular. For instance, Zarif had an approval rating of 76 percent, Rouhani’s was 75 percent, Khatami’s was 65 percent and Larijani’s was 45 percent. But now, popular approval of this group has fallen sharply.
With regard to “unpopularity” we see the same changes. For example, in 2016 Zarif’s disapproval rating was at seven percent, but this is now at 34 percent. The unpopularity of Rouhani has risen from 21 percent to 65 percent and Khatami’s disapproval rating that was 22 percent in that year is now 33 percent. But Ahmadinejad’s unpopularity has fallen from 39 percent to 28 percent.
These findings show that people are less enamored with those they liked at one time, and have changed their opinions of them because they have been disappointed in them. In 2015, 25 percent of the people considered the reformist camp “effective” and liked them, but this year this number has fallen to 11 percent. Meanwhile, the popularity of the principalists has not changed to a meaningful degree. Their popularity stood at 13 percent and now stands at 11 percent.
And all of this means what?
I believe that this year is not going to be an easy one for the reformists because the record of Rouhani’s administration and the record of reformist representatives in parliament has made them less popular and less trusted among the people.