Talk:Australian electoral system

I was shocked and appaled to discover that this important area was not covered, Australia one of the few countries in the world not to have a completely shit form of 'democracy' hadn't had one word said about it, not a single bloody word. Shame shame shame.

As you can tell I am very biased and partisan in this entry, but at least I think I put everything relevant to this website in fairly. Which is subjective of course. Now ideally what I would like to have on this page is the entire point of view backed up with appropriate evidence from www.aec.gov.au and others that 'PV or IRV is little better than plurality'. I am sure such a case exists and I have read them, but not one backed up with evidence from the Australian example in effect from 1920. Special:Contributions/144.135.254.196 12:42, 23 April 2005

I note that cases for and against particular voting systems are better prosecuted on their respective pages. Moreover, the analysis of why we have a duopoly is woefully misguided. The chief reason is electorates wasting the votes of approximately half of everyone. -- V.jackson (talk) 10:18, 2 July 2022 (UTC)

Agreed. This was written poorly and with a lot of bias. I have added some links and tried to remove bias. Would be nice if somebody could add some more details on the reasons why STV is causing two party domination. It should spell out why such issues are not expected in cardinal systems. --Dr. Edmonds (talk) 16:57, 2 July 2022 (UTC)

That would be IRV. STV, being a proportional representation method, doesn't seem to cause two-party domination unless limited by something else. In Australia's case, it seems that IRV's pull is greater than STV's push (as it were). As far as I know, neither academic research nor EM gives any definite answer to why two-party domination occurs with IRV, and whether it can be generalized to all majoritarian single-winner methods. Possible reasons for two-party domination are:

* Center squeeze (unstable IRV outcomes when there are three strong contenders, e.g. Burlington; see also Taagepera and Grofman p. 346). Condorcet doesn't have that.

* Favorite betrayal (IRV has this, Condorcet does too, usual cardinal methods don't)

* District magnitude (SNTV with s seats seems to produce s+1-party rule in Japan, https://www.jstor.org/stable/193914, but on the other hand, France with single-member district top-two runoff has four or so party rule)

* Natural packing/cracking effects (a party that has 10% support everywhere being given only a few seats because it's the outright favorite few places, e.g. the Liberal Democrats in the UK). This affects all single-member district methods unless they have top-up seats.

* Too few dimensions to issue space (analogous to Malta, and Taagepera and Grofman's https://doi.org/10.1111/j.1475-6765.1985.tb00130.x pdf: https://www.socsci.uci.edu/~bgrofman/42%20Grofman.%20Rethinking%20Duverger%27s%20Law..pdf)

* Some interplay of these effects (e.g. low dimensionality issue space kept low by center squeeze effects)

The only things that seem definite are: IRV can tether STV's otherwise multiparty rule, and single-member districts don't force two-party rule (counterexamples being runoff voting in France, approval in Greece). Even Plurality's two-party rule may sometimes be limited to a local scope (Canada, India). Kristomun (talk) 17:41, 3 July 2022 (UTC)