Tuesday, September 19, 2017

A likely sounding analysis

Noted at The Conversation, the effect of having an optional survey on same sex marriage:
Last week’s Essential had Yes leading 69-28 among the 62% who will definitely vote, and 59-31 among the overall sample. Yes supporters are more likely to vote than No supporters, more than compensating for lower turnout among the young.

Furthermore, as Peter Brent writes, if everyone had to vote in the plebiscite, people who were grumpy about being dragged to the polls for something they perceived as trivial would be likely to vote No. With optional voting, these people are likely to toss the voting material in the bin.

With such strong support for Yes, No’s only hope is to persuade people to vote about different issues, such as safe schools and political correctness. These issues have little relation with same sex marriage, but the No campaign will highlight them in an attempt to persuade people to vote on these issues.
So, a potential 10% discrepancy by having it as an optional postal survey.  On the other hand, is it possible that younger folk, being more mobile in where they live than older families, might be more likely to not get their survey letter?   That might offset things by a few percent, I would guess.

But the point remains - this is a hopelessly inaccurate way of gauging genuine population support for such a matter.

No comments: