There's been a spate of pre-election polls from Georgia lately, most of them not worth the paper they're written on. All have been released by various interests to demonstrate their electoral strength and have little to no basis in reality. The only consequence of their release has been to increase public skepticism of polls.
President Mikail Saakashvili's party (UPM) released poll a few weeks ago that was greeted with howls of derision, but for a lot of wrong reasons. Most people I talked to categorically refused to believe data publicly released by BCG National Research, the firm that polls for the president's party (a high ranking official in the government I met insisted that's different than the serving as president's pollster, which suggested a bit of disingenuousness on his part, but we'll not quibble here. Also, BCG is run by the wife of the head of the Central Election Commission, a particularly stinky connection in this part of the world). The poll showed the president with a comfortable lead. Big surprise.
There's absolutely nothing wrong with the president's party releasing a pre-election poll which shows him leading; that's what happens in normal countries. It's also normal to take it with a rock of salt it because it came from the president's party's pollster. On its own, however, the poll is hardly evidence of fraud or intent to influence the outcome of the election. Fundamentally, it's meaningless. I found it amusing that no asked why BCG interviewed 13,000 people when a sample of 1300 would have sufficed, or why they didn't release much of their methodology. As we've mentioned here repeatedly, disclosure is the most important step toward increasing public confidence in polls.
Anyway, the Saakashvili campaign finally stepped up to the plate and hired a real pollster, Greenberg Quinlan Rosner Research from Washington. Among many others, VP Jeremy Rosner polled for NS/NU (Yushchenko's party) in the 2007 parliamentary election in Ukraine. He's the pollster that presidents and ruling parties hire when they want to be taken seriously outside of their own countries, particularly in Washington.
Rosner operates on the typical western model of strategic polling: as the president's pollster, he is responsible for plotting out a roadmap to victory for his client, which, by the way, often means delivering really bad news. As a hired gun, Rosner has nothing to gain by pretty-ing up the numbers if they're bad; if Saakashvili loses, GQRR loses. Pollsters like to win, and you don't win by making things sound better for your client than they are. In fact, a dose of reality is what a lot of leaders need, especially those who tend to surround themselves with suck-ups.
GQRR conducted a survey of 1200 Georgians on behalf of the Saakashvili campaign between December 5 and 13th, with a margin of error of plus or minus 2.8 percentage points. Interestingly, ACT Research Ltd., a second-tier Georgian data collection firm and not BCG, collected the data (Rosner may have addressed this issue on the call, but I got cut off after a few questions).
If there was bad news in this poll, Rosner didn't share it on the conference call briefing he conducted for the media today. To be sure, it is the prerogative of and standard practice for the campaign to not release the entire survey and only talk about the parts that communicate their message. It's a strategic document. Why in the world would they show their hand to their opponents?
Rosner pointed out that the survey was conducted before the latest foolishness by Badri Patarkatsishvili which probably has a positive effect on his client's numbers. Even so, his poll showed Saakashvili with a better than two-to-one lead over his nearest rival (Levan Gachechiladze, at 19% to Saakashvili's 42%) among all residents and a 30 percentage point lead over Gachechiladze among a subset of 846 likely voters (46% to 16%). This is good news.
Why would anyone be surprised by these numbers? Saakashvili is an incumbent and enjoys all the benefits of incumbency, including a lavish array of administrative resources, his use and abuse of which have been well-documented this election. He's running against a cast of minor characters with no organization, no strategy other than buying votes (which, if credible, can be pretty effective) and minimal levels of public support. He exploited these strengths when he called an early election. Smart.
Now, if I was Saakashvili, I'd be a little concerned that despite all these considerable advantages, I wasn't above 50%, particularly given the weakness of the opposition. The president has taken some pretty serious hits in the past few months, particularly after the events of November 7th. There's a good chance, however, as Rosner pointed out, that the 16% undecided will break toward the incumbent, putting him above 50% and avoiding a run-off. If not, he stands to win easily in a match-up with either Levan Gachechiladze or Badri Patarkatsishvili.
Rosner also revealed that by a 44% to 30% margin, more Georgians believe the election will be free and fair. More than one-fourth (26%) don't know. Sure, a plurality think the election will be free and fair, but three-in-ten believing it won't be free and fair is a pretty substantial number, especially considering the election hasn't happened yet. I'd like to see those numbers crossed against those who are not supporting Saakashvili and against those who live in Tbilisi, where election violations have been most publicized. But let me point out, my objection here is a matter of interpretation, not a dispute over the accuracy of the numbers (which I completely believe).
I hope things are calmer than they were when I was in Georgia a few weeks ago. I hope the opposition comes up with a better strategy to address cases of election fraud than taking to the streets with mindless provocations and that Saakashvili learned a lesson from the overreaction on November 7th. Violence on the street would be the worst possible outcome. I'll also be curious to see what percentage of Georgians think the election will be free and fair after January 5th.