By Clay Shentrup with some contribution by Warren D. Smith
Contrary to the beliefs of many instant runoff voting advocates in non-IRV countries, IRV produces two-party domination. Minor parties in Australia are apparently well aware of this, and want to get rid of it.
In constitutional change, the Democrats will seek to provide for a bicameral parliamentary
system with a House of Representatives and a Senate, both elected by proportional representation [which IRV is
responsible to the people.
The Australian Greens will:
2.1 work to re-assert the authority of the Federal Parliament by:
Are these third parties saying they want to get rid of IRV and replace it with a different single-winner voting system such as range voting? No. These parties express few or no opinions about which single-winner voting systems are superor and inferior, although they probably mostly believe that IRV is superior to plurality voting.
They are saying that they want to get rid of IRV (instant runoff voting) and replace it with multi-winner PR elections.
Why? Probably because all the Australian third parties keep losing IRV elections, whereas they do win a small but significant fraction of PR seats. For example (I am writing this in 2008) the Australian Greens, as Australia's largest third party, have, in their entire history, won exactly one federal IRV seat. This was the (Cunningham) House seat won by Michael K. Organ, who served 2 years (2002-2004). The reason Organ managed to win this seat was that it was not a normal race consisting of the NatLib candidate, the Labour candidate, and some third-party candidates. In this particular race, there was no NatLib candidate. The NatLibs in their current vote recommendation recommend the Greens above Labour. Hence we expect that Organ in this election got most of the NatLib votes.
Similarly, as I write this in early 2008, the number of third-party-held Australian federal IRV seats is zero.
Perhaps stimulated by Britain's upcoming national referendum to decide whether to switch to IRV (the citizens of Britain rejected IRV by 68% to 32% margin on 5 May 2011), in October 2010 a nationwide professional telephone poll of 1202 random Australian adults was conducted on (among other things) this same question. (There was this difference, though: Britain already was using plurality voting and the question was whether to switch to IRV; whereas Australia was already using IRV which they call there "the alternative vote," and the question was whether to switch in the opposite direction to plurality.) The poll found that Australians would prefer plain-plurality voting versus the preferential (instant runoff) system they presently use to elect their House if forced to choose one. The poll's result was 57% to 37% (with 5% don't know/refuse).
Many believe that, despite IRV's flaws, it will serve as an intermediate step toward proportional representation, by which they typically mean Hare/Droop reweighted single transferable vote (STV-PR). The reason they say that is that STV-PR resembles IRV voting (in fact the two are the same in the single-winner case).
We say this is an unrealistic idea, especially in a country like the USA.
Why is IRV an unrealistic stepping stone to PR? Because IRV leads to 2-party domination. It did so in Australia in IRV seats and Fiji (and Malta) and Ireland in IRV seats despite the fact that Australia and Ireland also had strong third parties in non-IRV seats, and despite other factors conducive to third parties (like non-presidential, multiethnic/lingual). So in the USA, which lacks those third-party-genic factors, for sure there will still be massive 2-party domination under IRV voting.
Now the top two parties who are enjoying 2-party domination, are not interested in switching to PR! Many US cities adopted STV-PR, most famously New York and Cincinnati, and every single one abandoned it except for Cambridge MA. The two major parties both made heavy propaganda and political efforts to stop PR and those efforts succeeded. They would need to make much less impressive efforts to merely stop PR from being born, as opposed to stopping it once it already was going strong. So we can safely predict that, in the 2-party domination that IRV would continue to cause, PR would not happen in the USA.
The "stepping stone" idea won't work.
A better idea: Advocates of proportional representation, especially those in minor parties (or "third" parties in the U.S.), should instead consider supporting range voting (RV), as an intermediate step to reweighted range voting (RRV) or asset voting. These two PR voting systems are similar to range (indeed RRV and range voting are the same thing in the single-winner case). They both are simpler than STV-PR and have better properties (for example "monotonicity").
There is good reason – both theoretical and based on real data 2 – to believe that range voting (unlike IRV) is likely to erode 2-party domination (probably gradually). In all range voting exit poll studies so far, the third parties got into the same ballpark vote total as the major party candidates. That contrasts with almost every IRV election held so far, where the third parties are usually hugely far behind both major party candidates. As a rough approximation, the best 1% of historical performances by third-party candidates under IRV, will be approximately the same as the typical third party performance under range voting.
If reformers and third parties can bring about a switch to a better single-winner election method, with the goal of eventually transitioning to proportional representation, there's no telling how long that interim might be. Considering that it could last forever, it would be greatly to their benefit to choose the best single-winner system possible. In that case, the best option would be to get range voting implemented first, and then start working toward proportional representation.
For those who object to the characterization of Australia's third parties as being "anti-IRV", when in fact many of their members say they support IRV for single-winner elections, there is a case to be made that they actually do not support IRV—they just think they do! That is, if they don't like being shut out of multi-seat posts by IRV, why would they want to be shut out of single-seat posts by IRV?! The problem is that there is a widely held false belief that no voting method can prevent two-party domination in single-winner elections. This is false – history has shown that (non-"instant") top-2 runoff usually leads to more than two powerful parties. It's reasonable to presume that if third parties realized this, they wouldn't even support IRV for use in single-winner elections.
Return to main page