Dunkley byelection, page-447

  1. 14,237 Posts.
    lightbulb Created with Sketch. 5
    Yeah, nah... it ain't that simple, Peter...

    I was reading a blog post from Peter Brent the other day. He thought that the Libs would take the seat. That was his tip. But he also stated in the article that byelections are unpredictable.
    Brent in his calculations decided to follow Antony and use 1983 onwards but also excluded those section 44 byelections where the affected members ran again as candidates.

    https://insidestory.org.au/dunkleys-rorschach-test/
    So what can we say about Dunkley?
    Antony Green’s page is up, and I’ve followed his lead when calculating average swings by restricting the time period to 1983 onwards.
    But I’ve also excluded by-elections caused by section 44 of the Constitution — of which we had a slew around six years ago — because in all of them (or at least those with identifiable with two-party-preferred swings) the disqualified MPs ran again.

    These deserve their own category given that the absence of the personal votes of sitting MPs is the big driver of the difference between swings in opposition-held seats and government-held seats.

    That leaves twenty-three by-elections in the past forty-one years with two-party-preferred swings.

    In the ten opposition-held seats (including Aston and the low-profile Fadden last year) the average swing was an almost negligible 0.8 points to the opposition.

    Those caused by resignations by government MPs (eleven in total, the most recent in Groom in 2020) average to a much bigger number, 7.6 points to the opposition.
    And when they’re brought on by the death of a government MP — it’s a tiny sample of two (Aston 2001 and Canning 2015) — the swing is 5.5 points to the opposition.

    If we include that pair with the resignations we get 7.2 points to the opposition from thirteen events.(There were no opposition by-elections caused by death with two-party-preferred swings in that period.)
    So you might want to use that 5.5, which would see Labor retain the seat, or 7.2, which wouldn’t.
    Or you could slot in any other number, because another feature of by-elections is that they’re unpredictable.
 
arrow-down-2 Created with Sketch. arrow-down-2 Created with Sketch.