Methodologies

Q&A: Craig Charney, the World’s First (and Only) ‘Crisis Pollster’

By Dan Kedmey | Time Magazine | August 13, 2013

Craig Charney may be the only person on earth who can claim the job title of “crisis pollster.” In fact, no one had probably thought to put those two words together until 1997, when Charney, with the help of his polling team at Charney Research, began surveying people in the world’s worst conflict zones. He had already polled black South Africans in the darkest days of apartheid. In 2003, he conducted the first national survey of Afghanistan. His interviewers have even penetrated the anarchic tribal zones along the Afghanistan-Pakistan border. If being constantly exposed to countries in turmoil has made him cynical, he doesn’t show a trace of it. If anything, he smacks of an Enlightenment thinker who believes that tremendous forces of change can be unleashed anyplace, at anytime, as soon as the little guys find their voice — and he has the data to prove it. “We found we could do a lot of things that other people said were impossible,” he says of his polling work. TIME asked him to explain.

What exactly does a crisis pollster do?

We work in 45 countries from fast growing BRICS to full-fledged countries in flames. We did the first national surveys in Afghanistan, Cambodia and East Timor and we did the first scientific surveys on the border region between Afghanistan and Pakistan. We’re kind of specialists in tap-dancing around the mouth of the crocodile, but what people don’t understand is that we’re tap dancing in a safe spot behind the head where his mouth is propped open.

How did you become a crisis pollster?

I was a journalist in South Africa in the early 1980s, doing polling work with the marketing department of the Johannesburg Star. They asked me to do a poll of how whites would vote and how blacks would vote, if they ever had the right to vote. We called Nelson Mandela as the winner. I said the leader of a banned organization, a convicted terrorist, was the most popular person in the country. At that time we couldn’t show a picture of Nelson Mandela in the papers, so we showed him as an outline with a silhouette. In some ways that was even more effective, because it represented the shadow Mandela cast over South Africa in those days.

What was the reaction?

The authorities showed up at my door the next day to question me. Mandela noticed too. Ten years later he hired me as a pollster for his party.

What lesson did you take away from the poll in apartheid South Africa?

The most important being the most obvious, namely, that people in the developing world have an opinion and it matters more than ever before. We did the very first focus groups for Mandela. The fear at that point was that black South Africans would be afraid to speak up. It turned out the problem was not that they wouldn’t talk, but that they wouldn’t shut up. They were more than happy to speak. People — especially oppressed people — are more than happy to speak when it’s safe.

Even in a warzone like Afghanistan?

Even in Afghanistan, where they might not have an idea of what a poll is — we did the first poll ever conducted there in 2003 and 2004 for USAID. At the end of transcripts in several places, the standard conclusion was, “Salaam aleiykhum, thank you for answering our questions.” Several responses were, “Thank you for asking. No one had asked me for my opinion before.”

How do you make your subjects feel safe?

It depends on the situation. A survey operates on the principle of anonymity, that the interviewee’s identity will never be revealed. If the respondent is confident of that, then they’ll generally speak freely. What we try to do is trace the frontiers of what people feel they are willing to talk about and we stay within those boundaries. This is why it makes sense to have a crisis pollster. There are countries that have similarities in the structure and the problems they face.

How do you know when your subjects are being candid?

We interview them at home. We tabulate differences when people are interviewed alone, with family members, and in public places. Even when people are afraid to speak out, they’ll actually tell you they’re afraid. You have to stay outside of the red zone. This is not just a matter of getting good results; you don’t want to get them into trouble. We also compare the results to things we can verify, like election results. For instance, in Afghanistan in 2004 many experts thought women wouldn’t vote [in the presidential election]. We found they would — and they did. In 2009, the smart money was on Hezbollah to win Lebanon’s election. We were the only ones who said their pro-Western opponents would win — and they won.

You’ve also done polling work in fast-growing economies like China. How does it differ from polling work in conflict zones?

China is a pretty advanced sophisticated society and people have a pretty clear understanding of what surveys are. Other than talking about politics, it’s a normal research environment. The real lesson was the degree of openness that does exist in China — how much information flows. The late David Apter had this theory of modernization: The key thing about modernization is that information needs to flow in a society. The Chinese government recognized that information needs to flow much more than it did in the old top-down system. In this sense, it reminds me a lot of the other transitional countries we’ve worked in, where a process of change was underway, but not complete.

So you see a common process of change across all of these countries?

The thread that brings together the different parts of my work, whether its crisis countries, emerging markets and BRICS, is that public opinion matters now more than ever before.

Didn’t it always matter?

Seventy or 80 years ago, would an American business have cared that much about what was happening to interest rates in Shanghai? Would the future of foreign policy have depended on peasants in the Helmand province of Afghanistan or shopkeepers in the markets of Damascus? Put the question that way, and I think it answers itself.