Opinion poll reveals five favourites in Senatorial campaign

Five Senatorial candidates share a clear lead over the other ten according to the results of Jersey’s first ever online electoral opinion poll.

438 people participated in our poll, casting a total of 1860 votes across the fifteen candidates between 1pm on Thursday 29 September and midnight on Monday 3 October. The results show five candidates clearly ahead of the pack, with a marked drop to a pair of candidates battling over sixth position, and then a further gap before the remaining eight candidates come in.

Bookmakers’ odds favour the same top three as indicated by the Elect Jersey 2005 poll, albeit in a different order. At 5pm on Tuesday 4 October, Ahiers at St John was offering 2-1 against Senator Stuart Syvret to top the poll, 7-2 against Senator Terry Le Sueur and 5-1 against Freddie Cohen.

Matthew Robins, Co-chairman of Elect Jersey 2005 commented, “This poll can only be taken as an indication of the direction of public opinion in the Senatorial election. There are still two weeks of campaigning to go, and it’s entirely likely that the final order will be quite different, but a clear trend is emerging. Supporters of the top five will certainly not want to become complacent. With two weeks to go before polling day a lot can change.”

The poll was launched with an e-mail to the 15,000 users of Jersey Post’s free Jerseymail e-mail service. This produced a rapid initial response showing a pattern of voting which continued until the poll closed. In a response that suggests that the electorate is not impressed at the quality of the field, only 161 participants cast all six of their votes, and 88 (20% of the total) cast only two votes.

Matthew Robins added, “The results suggest that people are voting for individuals rather than parties. For example, only 42 of the 82 who voted for Senator Paul Le Claire also voted for his Centre Party colleague, Kevin Lewis. And every possible permutation of two candidates appeared. Independent thinking is alive and well amongst Jersey’s voters.”

Voting statistics

438 people voted.
There was one “spoilt paper” (voted for all candidates) which was excluded from the final count.
Only 161 voters (37%) used all 6 votes.
66 used 5 votes.
67 used 4 votes.
55 used 3 votes.
88 (20%) placed no more than two votes, of which:
- 43 only placed two votes, the most common combination being Terry Le Sueur and Freddie Cohen (8).
- 45 only placed one vote, the most common single votes being for Shenton (7) and Syvret (7).

Party voting

Only 46 of the 76 who voted for Deputy Geoff Southern also voted for Denise Carroll (16 voted for Denise Carroll plus others.)
Only 42 of the 82 who voted for Senator Paul Le Claire also voted for Kevin Lewis (25 voted for Kevin Lewis plus others.)
34 of the 82 who voted for Senator Paul Le Claire also voted for a JDA candidate.
23 of the 76 who voted for Deputy Geoff Southern also voted for a Centre Party candidate.

Other voting patterns

There was not a single candidate whose voters failed, between them, also to vote for all the other candidates. In other words, every possible permutation of two candidates occurred in at least one voter's selection.

Notes on the methodology

The poll opened at 1pm on Thursday 29 September and ran until midnight on Monday 3 October.

It took place entirely online using a well-established third party survey product.

It was "self-selecting", meaning that those who took part did so by their own choice. They were not selected to represent a balanced demographic mix. Such an approach would have been beyond the resources of Elect Jersey 2005. The results must therefore be seen as an indicator of how the candidates are being received by the electorate, and not treated as a reliable prediction of the election result.

The sample size (438 respondents) represents a significant proportion of the number of registered voters (now over 54,000.) By comparison, MORI's most recent monthly "Political Monitor" survey (August 2005) polled 2047 respondents for the whole UK population.

It was not impossible for the survey to be abused, but the survey web site does include safeguards to attempt to prevent people voting more than once. The data shows no evidence of mass voting patterns planned to influence the results in a particular way.

The launch to 15,000 Jerseymail users was deliberately planned to ensure that the initial response was not dominated by our web site’s regular users.

The voting patterns observed in the initial rush of responses following the mailing to Jerseymail users continued largely unchanged until the poll closed.