Exit Poll Under Fire

Opposition forces decry plans for an election day survey they say will favor Mikheil Saakashvili. From EurasiaNet. by Giorgi Lomsadze 2 January 2008 TBILISI | Plans to run an exit poll for Georgia's upcoming presidential elections and plebiscites are raising a fresh storm of political controversy in advance of the 5 January election.

Key opposition candidates have denounced the polling project as biased in favor of Mikheil Saakashvili, who stepped down as president in November to run for re-election, while some local observers fear that the survey may only add to existing tensions.

The poll, officially meant to act as a safeguard against election fraud, was commissioned by four television companies – the state-financed Georgian Public Broadcasting and the Rustavi-2, Mze and Achara TV private broadcasters. The opposition has routinely denounced Rustavi-2 and Mze for a pro-government bias in their news coverage.

Two prominent think tanks – the Caucasus Institute for Peace, Democracy and Development and the Georgian Foundation for Strategic and International Studies – have joined with the Georgian Institute of Public Affairs and government-funded Ilia Chavchavadze State University to manage the project. A slew of local polling companies will do the actual fieldwork and tabulation.

"We have no confidence in these exit polls," said presidential candidate Davit Gamkrelidze, leader of the New Rights Party, on 27 December. Speaking to reporters, Gamkrelidze alleged that officials will manipulate the exit polls to try to make the election results appear legitimate. Gamkrelidze, along with fellow candidates Levan Gachechiladze and Shalva Natelashvili, have charged that the government will attempt to rig the outcome in favor of Saakashvili, whose crackdown on opposition demonstrations in early November tarnished his image as a reformer.

At a 29 December rally in Tbilisi staged by the nine-party coalition backing Gachechiladze, some participants, quizzing a reporter about Washington's view of the election, pushed for international observers to stop the polls. "The exit polls are a trick being used by the government. That's it. Everyone here knows the danger," argued one Gachechiladze supporter, wearing the white scarf associated with the campaign.

In response to such criticism, the government has claimed that opposition leaders are not willing to cooperate to hold a fair election, but instead are focusing on stirring up trouble after the election. "Rather than preparing for the election, the opposition is getting ready for 6 January … this is deeply troubling," Acting President Nino Burjanadze said on 26 December in televised remarks.

MONITORING THE MONITORS

Trying to allay tensions, the group overseeing the exit poll has invited critics to monitor its execution. "I call on political parties who have doubts about the exit polls to send their representatives and they will be allowed to monitor the whole process," Temur Iakobashvili, president of the Georgian Foundation for Strategic and International Studies, told a news conference on 30 December.

Covering all electoral districts, the exit polls will be conducted at 228 of some 4,000 polling stations. At least two pollsters will stand outside each polling station; one will keep a headcount of exiting voters and the other will ask every ninth person to say for whom he or she voted.

In an effort to reduce the non-response rate, voters will be given an opportunity to answer confidentially by marking the candidate's number on a piece of paper designed to resemble the ballot and placing it in a box.

Control groups drawn from polling companies will check the performance of pollsters. Results are to be made public by the end of Saturday.

Mamuka Nadareishvili, a statistician at the Social Assistance and Employment Agency who worked on exit polls for the 2003 parliamentary and 2004 presidential elections, has been hired by the Foundation for Strategic and International Studies to design the poll's methodology and to coordinate its implementation.

Nadareishvili argues that the exit poll is a powerful tool against the "dead souls" trick – the use of names of deceased voters to cast votes – as it will record the exact turnout at selected polling stations.

If the non-response rate does not exceed previous ranges of 20 to 30 percent of respondents and if the race is not too close to call, the poll should give an accurate account of how voters actually voted, Nadareishvili forecast. The poll's survey design has a margin of error of 2 percentage points.

But the poll may or may not reflect the official results, Nadareishvili added. For the 2003 parliamentary elections, the difference between exit poll results and official results fueled opposition charges that the election had been rigged.

"In theory, I can't rule out a situation where all the non-respondents vote for a single candidate and that could dramatically change the entire picture. In reality, however, that is extremely unlikely," Nadareishvili said.

‘NOTORIOUSLY UNRELIABLE’

One local polling practitioner differs. "Exit polls are notoriously unreliable," said Hans Gutbrod, regional director for the Caucasus Research Resource Centers. Even if the pollsters get the methodology right, significant deviations from the actual vote count are possible, he noted. That track record holds particularly true in Georgia, which has little experience with exit polls and little verified information on past voting patterns, he noted.

Gutbrod, whose center oversees a Caucasus-wide annual opinion survey, argues that opinion polls should be conducted repeatedly to get a sense of region-specific voter behavior. If a country is relatively new to exit polls, he contends, "you have a [die] that is not properly balanced. The problem is that you may randomly select one polling station in a district, but you don't know if the sample is representative of the entire district."

Given such shortcomings, the exit poll results could differ sharply from the official tally and fuel further controversy over the outcome, Gutbrod worries.

Signs of such potential trouble are already in place. Almost all of the candidates have carried out their own opinion surveys, and each among them claims to be in the lead. Opposition media tend to run or commission polls that show opposition coalition candidate Levan Gachechiladze skimming past Saakashvili by several percentage points. A 31 December poll run by the pro-opposition newspaper Rezonansi named Gachechiladze as the top choice for 22.4 percent of 801 respondents; Saakashvili trailed at 17.6 percent.

A controversial poll commissioned by the United National Movement Party and published in early December, however, put Saakashvili at 54.5 percent of the vote, with Gachechiladze at a distant 15.5 percent. The poll was run by a company headed by the wife of Central Election Commission Chairman Levan Tarkhnishvili, a fact that the opposition has cited to reject its results.

Gachechiladze, who is viewed by many as the strongest challenger to the former president, has pledged to fight the official election results should Saakashvili be reelected.