During the lead-up to the 2013 Iranian presidential election, the earliest signs of Hassan Rouhani’s initially unanticipated victory came from a surprising source: high-quality polling conducted by Iranian pollsters. In the runup to the presidential election in May 2017, all signs point to a greater availability of Iranian polling data than in the past.
ESOMAR World Research, the polling industry’s international trade association, lists a dozen Iranian members on its website, a number comparable to other emerging market economies, like the Philippines. The Iranian Marketing Research Association has 24 institutional and individual members listed on its website. With the opportunity presented by the next presidential campaign in Iran, one can expect new players to enter the market.
The first question many ask, given the nature of the Islamic Republic’s political system, is whether public opinion surveys conducted there can be considered valid. Modern Iranian political polling goes back to the late 1980s, and today high-quality polling can be conducted almost anywhere in the world, thanks to universal cellphone coverage and advances in computer-assisted survey administration. Pollsters, including Western organizations like the Pew Research Center, conduct reliable polls in China, Russia and locations in the Middle East. D3 Systems, a US contracting firm, was even able to conduct surveys in Iraq and Afghanistan during intense periods of combat there.
Iranian researchers are aware of the latest advances in survey methodology. In 2001, the Ministry of Culture and Islamic Guidance commissioned a poll on some controversial issues in Iranian society, among them political rights, social trust and religious behavior. Instead of asking respondents about their own personal views, the survey asked how they thought a majority of the population might respond. What is notable here is that this style of questioning is also preferred by Western pollsters for research on sensitive topics. That a ministry was involved in the 2001 survey suggests official approval for this type of research. Important global social science datasets, like the World Values Survey, also collect data in Iran.
The truth is, good survey research does come out of Iran and similar countries on a regular basis, and the data are used reliably by social scientists and political analysts. Nonetheless, some Iranian polls do fall short of international methodological standards, and the debate over survey quality is ongoing in English and Persian media.
After the contentious 2009 presidential election, supporters on both sides in the West cited public opinion surveys to argue that either Mahmoud Ahmadinejad or Mir Hossein Mousavi was the “real winner” of the election. Jon Cohen, who ran polling at The Washington Post at that time, found that some pro-Mousavi surveys had been conducted several months before the election and had high numbers of undecided voters. Those two factors combined are sure signs of unreliable data.
So what guidelines can observers use to get the best out of Iranian polling data ahead of the May 2017 presidential vote?
First, when encountering any poll, look for the disclosure of basic information about the survey and its methodology. Ignore any poll without it. The American Association for Public Opinion Research (AAPOR) publishes guidelines outlining minimum disclosure requirements that are influential worldwide. The AAPOR standard requires pollsters to publish essential information on the survey, such as how it was conducted, identity of respondents (e.g., likely voters, adults) and how they were selected, and the margin of error if applicable. Better disclosure means better surveys. With such information provided, not only are data consumers able to judge survey quality, but the ethics of disclosure help hold pollsters accountable to their colleagues in the research profession.
Western-based pollsters working in Iran generally abide by AAPOR-style standards, and for this reason their work can be considered reliable. These include Hossein Ghazian of IPOS, Ebrahim Mohseni of the University of Tehran Center for Public Opinion Research and Steven Kull of the University of Maryland. As Iran's domestic polling industry advances, more in-country researchers will also conform to similar standards.
Second, beware of polls that are not representative of the nationwide electorate. Media outlets based in Tehran sometimes conduct surveys only in Tehran province or the city of Tehran. Tabnak, Mehr news agency and other outlets released several such polls in 2013. That year, Tehran province represented less than 9% of the Iranian electorate and could not be considered a good national proxy by itself. When encountering such polls, ask whether the survey is representative of the entire electorate or just one part of the country.
Third, be alert to nonscientific online polls, mostly “conducted” by news websites and only among visitors to its pages. These “surveys” do not use representative sampling and should be ignored for that reason. Ask whether the survey explicitly attempts to be representative of the entire electorate.
In terms of these simple guidelines, the 2013 polls performed remarkably well. First, they showed Rouhani’s momentum in the final two weeks of campaigning: a move from single digits to nearly 40% of the vote. Such big swings in the polls are rare in the West, but they are not unusual in emerging market countries like Iran, especially ones with relatively weak, formal political party systems.
Second, the polls correctly identified Rouhani’s main rival: Tehran Mayor Mohammad Bagher Ghalibaf, then largely unknown in the West. Ghalibaf came in second, ahead of such buzzworthy candidates as former chief nuclear negotiator Saeed Jalili and Ali Akbar Velayati, the foreign policy adviser to Supreme Leader Ayatollah Ali Khamenei.
This level of accuracy represented a big success for Iranian pollsters, and it is a positive development for all observers of Iranian politics. Good polling shows that Iran has not been excluded from the recent global trend toward a data-driven better understanding of politics. As such, 2017 may be the year of the first Iranian Nate Silver.