Poll Shows Bainimarama Well Ahead
The poll, which will be conducted weekly until the elections, comprised a sample of 600 eligible voters, 300 from the Central Division, 200 from the West and 100 from the North. Those interviewed were "ordinary people passing through bus stations to get a range of opinion from both urban commuters and from people travelling into town from rural areas."
There's little doubt the methodology and results of the poll will be questioned by Bainimarama's opponents but the results are remarkably close to those to those of the 2011 Tebbutt Research poll conducted for the Australian think-tank, the Lowy Institute, that found 66% of Fijians thought that the Prime Minister was doing a good job.
I have had considerable experience in drafting, conducting and evaluating surveys in Fiji and think the CJP methodology adequate to the task except that I would have preferred a somewhat larger sample size and a margin of error. Not that in this first poll a margin of error would have meant anything. My main reservation is that survey results in Fiji are influenced by the age, sex and ethnicity, not just of those interviewed but also by those conducting the interviews. This was found to be especially true on "sensitive" issues. My research found that men and women answers on rape differed with the sex of the interviewer, and iTaueki, Indo-Fijian and Other Ethnic answers on ethnic balance in the RFMF differed with the ethnicity of the interviewer.
CJP may not be free to show the poll results by ethnicity. Government seems to have ruled that national statistics should not reveal ethnicity. This means that airport arrivals and departures and census results no longer reveal ethnicity, making them much less useful for analytical and monitoring purposes. But this should not prevent CJP from asking the age, sex and ethnicity of those interviewed and then conducting an internal check to see whether they affected the results.
Not to be outdone, Fiji Today publisher and Government critic "Peter Firkins" conducted his own admittedly unscientific poll. His business takes him around Fiji. He asked 432 people during February "how Fiji is going?" Some 47% supported "Frank and what he is doing"; 8% supported the FLP, 34% SODELPA, 4% another party, 6% were not willing to say and 3% were not interested in politics.
He did not reveal the ethnicity behind the figures but they would obviously be influenced by the proportion of i'Taukei and Indo-Fijians in his sample.
He added, "If Qarase was still the head of the SDL he would have significantly more support in what appeares to be a form of sympathy vote. The younger new voters would vote for another party if it could capture their imagination but will vote for Frank rather than the 'Old' parties."