Varying political party survey results on the eve of polling day in Estonia may be viewed in a positive light by Conservative People's Party of Estonia (EKRE) supporters and by Reform Party supporters alike, according to political scientist Tõnis Saarts.
No fewer than four research companies have issued polls this week, the week when advance voting online and at polling stations began.
Regulars Kantar Emor, Turu-uuringute and Norstat were joined by RAIT Faktum & Ariko this week, and while the results of all four companies' questionnaires, since they are conducted using different methodologies and in different windows of time, need not be identical, there is a particular discrepancy in the rating given to EKRE, from polling company to polling company.
This range runs from 14 percent to 25 percent, a margin far larger than the usual margins of error claimed by pollsters.
Appearing on ETV current affairs show "Ringvaade" Thursday, Saarts said these variations were causing confusion among the public.
"If evaluating these studies with a sociologist's hat on, I would say that it might seem especially confusing to the average voter, though much depends on the exact time-frame in which these polls were conducted; some events can affect the result, as can the number of people polled, i.e. the number of voters, how they were polled, i.e. how many online, how many face-to-face and how many over the phone, and also how questions were worded. Seemingly minor aspects like these can come into play," Saarts said.
Nonetheless, the variation on EKRE's reported rating can help both that party and the Reform Party (EKRE's de facto anathema-ed.).
"Paradoxically, I would say that this current flurry, in which some surveys show that EKRE and the Reform Party are neck-and-neck, while others state that EKRE is actually ahead, etc., is in some ways certainly good news for the campaign champions and offices, both of EKRE and the Reform Party."
"Since earlier on, a month or a few weeks ago, there had been more talk of the Reform Party definitely triumphing, with EKRE getting about 20 percent and the matter thus being settled, now the voters of both parties are being activated, along the lines of 'our party may win or, on the contrary, lose instead,'" Saarts went on.
EKRE and Reform supporters are generally the most loyal of all from all political parties, he added.
The most accurate polls are those conducted closest to polling day, Sunday, he added. Polling face-to-face also provides a clearer picture, he argues.
"I would generally have a little more confidence in those surveys which use face-to-face interviews as well, rather than just internet polls. What any voter looking at the survey results should take into account is that practically all of them have a margin of error of plus or minus two percent," he added.
Paradoxically, despite polls nearest Sunday being likely most reliable, Saarts added that longer term trends over the past month should be taken as the rule of thumb on how a party might be expected to poll on the day, rather than changes on a weekly, or now, even daily, basis.
"Should someone spot a difference between two parties of only one percent or less, or the rise or fall of one party support that is within this margin of error, then I do not recommend taking the results too seriously; rather, one should take a look at the broader trends - is it clear that some party or another has seen its support rising over a period of time, or falling in recent months," Saarts concluded.
Polling day is Sunday, March 5, with polling stations open from 9.00 a.m. to 8 p.m. e-voting closes at 8 p.m. on Saturday, at the same time as the advance voting period overall. On Friday and Saturday, all polling stations are also open, from noon to 8 p.m. Voters can cast and re-cast their e-vote as often as they wish, and can annul their e-vote altogether by voting on paper on the Sunday.
Editor: Andrew Whyte, Merili Nael
Source: 'Ringvaade', interviewer: Marko Reikop.