Trying to made sense of the published results is not easy. The most important thing we needed to know is how those polled intend to vote on 14 November. Not how well they thought the various leaders had performed in their respective positions. This is of some interest for PM Bainimarama and for SODELPA, because of the tussle between its two leaders, Ro Teimumu and Sitiveni Rabuka, but for the other leaders it was less important.
Responsibility for the faults that makes understanding difficult lies primarily with the Times journalists but Tebbutt did not help matters by asking this question for all the leaders.
The Times published two articles on the poll. One told us that 68% would vote for Bainimarama, 24% for Rabuka, 7% for NFP's Biman Prasad, and 2% or less for the others as PM if an election were held tomorrow. But the people do not vote for the PM. They vote for MPs and parties. It is the winning party that elects the PM, not the voters.
The journalist further confused the results by switching from "which leader people would choose to vote for" to "which you would prefer as PM."
Bainimarama was "the popular choice" (another new term) for the two major ethic groups, 93% of Indo-Fijians and 53% of Taukei, compared with Rabuka's 39%. Rabuka's support among Indo-Fijians was not stated.
Bainimara was also the "top choice" (sic!) in all administrative divisions: Western 75%, Northern 68% and Central 61%. Rabuka's respective figures were 18%, 28% and 29%, and Prasad's 5%, 2% and 7% .
Support drop not "significant"
The other Times article was headed "Support drops from 82pc." The "drop" —of 3%— was statistically insignificant because it could have been caused by sampling error. A more appropriate heading would have been "Support substantially unchanged."
The article dealt with the leaders' perceived performance in their present positions (i.e., as PM and leaders of the different parties) on a five point scale from very good to poor. Bainimarama scored 79% (presumably "a very good job" and "a good job") as PM, a statistically insignificance drop of 3% since the last Tebbutt poll in May. His disapproval rating (presumably "not a very good job" and/or "a poor job") remained unchanged at 5%, the lowest of all the political leaders performing in their own roles.
SODELPA's Ro Teimumu had an approval rating of 57%, 9% down from May, and SODELPA's present leader Rabuka an approval rating of 49%, down from 58% in May, and a disapproval rating of 21%, up 6% since May. The SODELPA ratings are complicated because Ro Teimumu had earlier said she was standing down from politics, Rabuka's court case could still see him being unable to stand, and internal feuding.
NFP's Prasad approval rating dropped from 51 to 36%, and his disapproval rating increased to 24% compared with 14% in May.
An earlier Razor poll indicated that HOPE is the only other party that stands a chance of returning a MP to Parliament. Its leader Tupou Draunidalo's Tebbutt poll approval rate was 30% and her disapproval rate 26%.
I doubt that much can be read into how these leaders, other than Bainimarama, performed in their present position, and see little purpose in asking the question.
My plea to The Fiji Times (meant to be helpful): Keep the wording of poll questions consistent, report all results for each variable, check that you have reported all the results, and consider using a table to summarise some of them.
Court case against RabukaRabuka may still not be able to stand for election. The FICAC has appealed the Suva magistrate ruling that it had "not been established beyond reasonable doubt that SODELPA was a registered party and Rabuka a party official" (a ruling I found quite strange.)
The case will now be heard in the High Court by Chief Justice Anthony Gates. The appellant (FICAC) must file written submissions by Friday 2 November, the respondent (Rabuka) by Tuesday 6 November, there will be an oral hearing on Thursday 8 November, and Justice Gates will deliver his judgment at 3pm Monday 12 November — two days before the election, but after many people will have voted.