Prime Minister Malcolm Turnbull is selling the postal vote on same-sex marriage as a chance for "all Australians to have their say" on the issue.
Subscribe now for unlimited access.
$0/
(min cost $0)
or signup to continue reading
"It's an important question. Everyone will get an opportunity to have their say," he said on Thursday.
There is a lot riding on the postal plebiscite. The government says unless there is a majority "yes" vote, it will not allow a vote in parliament.
But will the non-compulsory $122 million vote accurately reflect what Australians think about marriage equality?
Australia's leading pollsters have rubbished the postal vote approach, describing it as a "complete waste of money" and the "worst type of research, apart from robopolls".
They say if the aim was working out what Australians think on the issue, a scientifically sampled poll would be more accurate and could be done for a fraction of the cost.
Former Fairfax pollster John Stirton says postal surveys "went out with the dinosaurs" because they were both expensive and difficult to get an accurate reading of what people thought.
Mr Stirton, who now works as a research consultant, said a high quality survey of 2000 people (using live landline and mobile interviews, supplemented with online polling for tricky-to-reach demographics) would cost around $20,000.
A "gold standard super survey," that doorknocked 15,000 to 20,000 people (in every demographic), face-to-face would cost in the ballpark of $1 million.
But Mr Stirton said the results were likely to be the same, given the consistency of previous opinion polling on same-sex marriage, which puts support versus opposition at about two to one.
"You're more or less likely to come up with the same numbers, which are likely to be more accurate than the postal vote."
Essential Media Communications executive director Peter Lewis panned the postal option, saying only a robopoll would be a worse approach.
"If you got a whole bunch of researchers in a room and asked them to come up with the worst type of research you might come up with this [a postal vote]."
Mr Lewis said the non-compulsory poll would result in a self-selecting sample, as it's likely only the people that cared deeply either way would respond.
He added the postal vote method was a "high barrier to involvement". "It's making you have to physically do something and take it to a post box."
Mr Lewis also pointed to a further problem: the postal vote method meant older voters were more likely to respond, while same-sex marriage has more support among younger people.
"It's using a single platform that a lot of the target audience don't use."
Roy Morgan Research executive chairman Gary Morgan described the postal vote as a "complete waste of money".
Mr Morgan, whose company polls 1000 people every week, said people could be polled in each state and territory for about $3500.
"I can get you the result in 24 hours," he said, adding the $122 million postal vote was "not funny. It's just ridiculous."
Ipsos director Jessica Elgood, who runs the Fairfax/Ipsos poll said a survey with a sample size of 1400 would give a reliable picture across all states and territories, adding increasing the number to 2000 or 3000 would provide a "very robust take".
But, she cautioned there was also a democratic element to the postal vote, in terms of people having the opportunity to have their say.
"[A poll] does not carry the democratic stamp of approval."
Mr Stirton agreed even though a scientifically sampled poll would be more accurate than the voluntary postal vote, it would not have the same sense of legitimacy.
