Fearing the bandwagon effect: Georgia’s opinion poll wars

TBILISI, DFWatch – A new opinion poll in Georgia surprisingly showed that a new opposition movement has little support. Was it a sensational, true, finding, or did the pollster “cook” the survey?

The election campaign is well underway in Georgia, and apart from the ground rules for campaigning and campaign finance rules, another battle field emerged this week: Poll wars.

Already last November, the polls differed greatly, one showing President Saakashvili relatively secure in power, while another showed an almost even support base between him and the opposition bloc now known as Georgian Dream.

On the one hand of the war are opinion polls done by the foreign branches of two U.S. political parties, the International Republican Institute (IRI) and the National Democratic Institute (NDI).

These are controversial; every new poll is usually followed by a discussion in which there is significant distrust and suspicion towards the method and the way the results are presented.

When Bidzina Ivanishvili, the businessman who in 2008 had a higher approval rating (70%) than President Saakashvili (63%) started an opposition movement, it was widely perceived as a close to even match. One poll last November showed him almost neck and neck with the ruling party, 32 against 36 per cent.

But in NDI’s latest poll, Ivanishvili’s support is only 10 per cent, and the ruling National Movement received 47. Has the government-friendly TV channels’ inquisition of “the Oligarch’s” supposed Russian links undermined his support?

Technically, the result is only a rumor, because the NDI never publishes these polls. The results are leaked.

Usually, the NDI also doesn’t comment the poll results either, but this time local NDI head Luis Navarro was dragged into opposition-leaning Kavkazia TV’s studio and questioned about the poll and what his organization’s role is in Georgian society. One of the points of dispute between Navarro and program host David Akubardia was the NDI’s practice of not publishing their polls, but confidentially sharing parts of the poll results with political parties in a kind of order of rank, starting with the ruling National Movement party and moving down the list according to significance.

Within Ivanishvili’s opposition movement Georgian Dream, the number two party but almost insignificant if going by the latest poll, there is a strong distrust towards the NDI’s findings. The movement Friday published an open letter addressed to the U.S.’ ambassador John Bass, asking both the NDI and the IRI to refrain from conducting any more polls until after the election for parliament in October, apparently fearing the so-called “bandwagon effect”, which means that the public it will gravitate towards the party seeming to be winning in the polls. Bass answered that the NDI and IRI are not controlled by the U.S. government but carries out valuable work and he trusts their professionalism.

The pollsters might actually need Bass’ support. The polling industry in North America is basically self-regulating and there are no government-imposed standards of practice. What could have been a check is the companies’ fear of a tarnished reputation, but even in the West it happens frequently that polling companies ask atrocious question that just happen to produce results in line with the client’s position, what is called “advocacy polls”, according to Steven Brown at the Laurier Institute for the Study of Public Opinion and Policy, Waterloo, Ontario, an institution which monitors how polling companies are working. It’s not necessary to directly fabricate results to eschew the answers.

As to why a pollster would risk its credibility, most people are not very sophisticated around survey methodology, the threshold for being discredited is reasonably high and therefore often worth the risk if the return is high.

NDI’s poll was conducted by Caucasus Research Resource Centers. According to Tamar Khoshtaria, one of the researchers responsible, the poll was done by going from door to door in all regions of the country between February 22 and March 5. She says the sample was representative and reflects different social strata. Fifty interviewers participated, and they were all trained beforehand.

-We will try to be more active in the future (laughs).

-I’m sorry, I don’t have that information, and I don’t think it is public, says Khoshtaria. She is adamant that there was no agenda behind the CRRC’s poll for the NDI.

-Our research organization is not pro-governmental or pro-oppositional. We try to have high standards. I can’t talk about other research organizations, because I don’t know how they work, she says.

The main criticism against the NDI poll is that there is a fear among people of saying their opinion. The CRRC did not use any particular method to alleviate that fear. The competitor Institute of Social Studies and Analysis (ISSA), on the other hand, did; when it found Saakashvili’s and Ivanishvili’s competing blocs pitted against each other 36 to 32 per cent last November.

-Sometimes it’s real, sometimes not real, but the fear is that if you openly support an opposition party then you might lose your job, or your family member might, and so on. This works very well and people really can not express their views openly, says Teona Mataradze, one of the researchers at ISSA.

-This is also expressed in the fact that number who refused to answer or don’t know is very high in the last [NDI] survey. Of course we can not say that all those people who refuse to answer support an opposition parties. Some part of the population really are undecided. Still there is more than half a year till the election. But I still think that there is a very clear factor of fear. And it’s increasing especially before the election, because in the media and everywhere there is discussion.

-Also, the percentage who refuse any kind of interview is becoming higher. People are either fed up with this, or they are just so afraid that they don’t want to give any kind of interview, she says.

In November 2011, a poll by ISSA showed the National Movement leading with 36 per cent to Ivanishvili’s 32. A poll done by the IRI around the same time showed a different results: 42 per cent for the National Movement and 18 for Ivanishvili.

-I don’t know. It depends on sample size and on guaranteeing the anonymity of the questionnaires. Sometimes we are using envelopes in order to ask the respondents to express their position. It sometimes helps them to be more open and more neutral. I think that makes the difference, Mataradze tells DF Watch by phone.

Andrea Perrella, the Director of Laurier Institute, explains that there is a constant debate among the polling industry about how to best execute and administer a survey in order to draw a natural sample of the population. In a Georgian context, there might be an issue with whether pollsters are measuring the opinions of those who have disengaged from politics.

-If there is some cynicism or some disengagement in a society, then the question is when conducting a poll: Who are you surveying? Are you surveying those people who are engaged, and those who are not engaged are choosing not to participate? Therefore their voices and their opinions are not being reflected in the survey, Dr Perrella says.

Going from door to door is generally a better way of getting a representative sample than phone calls, he explains. But there are also problems with choosing neighborhoods, communities, and the method tends to be expensive. In the West, people are more reluctant to not answer if you knock on someone’s door than call them. But that might not be true to the same degree in the former Soviet Union.