In a week where opinion polls suggest both Hillary Clinton and Donald Trump are leading the race to become the next US president, experts in American politics give their take on possible reasons behind the fluctuation.
As the race for the White House nears its end, various independent opinion polls have suggested that both candidates have a sniff of winning the US presidency when voting booths officially open on November 8.
US affairs lecturer at Melbourne University, Emma Shortis, said knowing the exact reason behind such a fluctuation was "very hard".
"Four years ago we had polls saying that Obama was going to lose the election, and he didn’t."
She suggested that undecided voters swinging back and forth may be to blame.
"It could be as simple as the nature of the poling and how it's conducted. But I do believe that a measure of undecided voters swinging towards one candidate or the other may be causing it," she told SBS.
"Things (in the race) are changing so quickly and I feel that the dramatic revelations that we just keep seeing are having an effect on the campaign."
Ms Shortis said that a factor over the past six months has been traditional Republican voters wavering in their support for their party.
Poll averages a better predictor
Despite holding weight with the US electorate, single opinion polls are not necessarily the best way to predict a race leader, according to Associate Professor Brendon O'Connor from the US Studies Centre at Sydney University.
He suggested looking at an average of figures, accumulated from several polls, as a more accurate indicator.
"If there are 22 polls in a relevant period of time, and one says Trump is going to win, and the other 21 say Hillary Clinton’s going to win – why would you focus on the one poll?
"Look at what the average is, look at what the trend is on a composite of polls, rather than getting concerned about one poll," he said.
Prof O' Connor said such a close race in pre-election polling was not exclusive to the current race, pointing to the fluctuating figures that appeared before the 2004 and 2012 US elections.
"With Gore and Bush in 2000, the polls were all over the place and there were very different predictions. Even with Barack Obama (in 2012), there were a number of polls that said that (Mitt) Romney would win all year long," he said.
"Four years ago we had polls saying that Obama was going to lose the election, and he didn’t. The majority of polls said he was going to win and I was quite confident to be able to predict the election result a week out from the election - that Obama would win because of the poll averages."
Trump swing not enough
The Trump camp took a major hit in opinion polls in early October after a video surfaced showing the New York billionaire making sexist remarks about women.
A poll by Monmouth University during that period saw Clinton open up a 12-point lead.
But the tide began to turn as polls taken in November suggested Trump had gained on Clinton following the FBI's announcement that it would investigate her use of a private email server.
Despite this "totally credible" late surge by Trump, Prof O'Connor said Clinton remained the firm favourite due to the demographic breakdown of the US.
"Trump is going to get a lot of votes from white men and elderly whites," he said.
"But, there’s enough women, young people, blacks, Hispanics to stop Trump from getting the presidency. He has alienated too many different groups that the Republicans need (to win).
"In this tightening race, the enthusiasm for both candidates is fairly low."