You are not permitted to download, save or email this image. Visit image gallery to purchase the image.
Polling has come under fire for the possibility it may stop people from voting if their chosen candidate is either well ahead, in an apparently unassailable lead, or well behind, with no hope of winning.
There has also been a suggestion it could be "self-fulfilling".
But Dr Chris Rudd, of the University of Otago, said yesterday, from an academic point of view, there was "no clear-cut evidence that polls influence turnout, or have a bandwagon or underdog effect".
The Otago Daily Times, and other New Zealand newspapers, published the results of pre-election polls during this election, as has been done in the past.
The ODT published two polls of the city, one on September 26, and one on October 6, three days before voting closed.
The first poll showed the eventual winner of the Dunedin mayoralty, Dave Cull, ahead with 45.9% to incumbent Peter Chin's 37.4%.
In the second poll, Mr Cull's support stood at 54.2% - 5.9% less than he eventually received on election night.
Mr Chin's support in the second poll was at 31% - 0.4% less than Saturday's vote.
Federated Farmers last week criticised the publication of polls in Auckland and Christchurch - both of which correctly predicted the winners - before the elections closed.
President Don Nicolson asked why local elections were treated differently from general elections.
"I think it's time to look at reforming the time period voters have to return their papers."
Mr Nicolson said the Electoral Act prohibited the publication or broadcast of election-related material after 6pm on the day before a general election, until booths closed.
"It includes a complete ban on election-day opinion polling, for the simple reason it could influence voters.
"So releasing what to all intents and purposes are exit polls, could skew local council voting, because it may lead to people choosing not to vote on the basis their favoured candidate is either in pole position or an also-ran."
Mayoral candidate Lee Vandervis also criticised the polls, and said after the September 26 publication they were illegal in some countries because they potentially gave the media opportunities of skewing voting behaviour.
Mr Vandervis said he believed the increasing ease of producing surveys with new technology would probably see a trend to more surveys in the future.
He saw the "tendency for them to be self-fulfilling".
"I worry about them pre-election ..."
But Dr Rudd said there had been "heaps" of studies on the issue.
Those included experiments using mock polls to study behaviour, and there was no clear evidence of a "bandwagon effect", or of people voting for an underdog.
As Mr Cull received less support than the ODT poll predicted, there had been no "bandwagon" effect.
Dr Rudd noted Mr Vandervis' support had gone up after the polling showed him with little chance of winning.
"It didn't affect him."
Of the ODT poll, Dr Rudd said the differences between the poll and the actual outcome were "remarkably small" and it would be odd if there were no changes in voting between the poll and polling day.