[EM] Why does IRV but not delayed top-two runoff lead to 2-party domination?

[Juho]

The web page listed some interesting factors that may have effect but  
I think it didn't provide a complete plausible explanation to the two  
party domination questions and the role of IRV and top-two runoff in  
this.

 > Why do those two methods, which seem strategically quite similar,  
lead to such different results?

I think the methods are behaviourally closer to each others than what  
the web page said. There are also other reasons behind two party  
domination. I'll try to address some of them below.

 > 1. Different strategy calculations by voters under the two systems.

I didn't find convincing evidence here. There are many strategies the  
voters could try, sensible and less sensible. Some of them, like part  
of supporters of one candidate voting another one, are very hard to  
control successfully in large public elections.

 > 2. ... last round ... attention from the media

This may have some impact but I don't expect this influence to so big  
that it would clearly differentiate the two methods.

Some other observations:

The number of analysed IRV countries was small and many of them have  
their own peculiarities. A more detailed analysis would be needed to  
make conclusions from them.

The page indicated that the listed top-two runoff countries use that  
method for some presidential elections. I guess most of them use  
other methods to elect the (multi-party) parliament.

The parliaments (and their "multi-party members") are often elected  
in multi-winner elections, not single-winner (like IRV and top-two  
runoff). I'd say that the division to countries using single-winner  
districts vs. multi-winner districts (+ a proportional method) is a  
more important explanation to why some countries become two-party  
systems than the election method that is used in the single winner  
(president or parliament) elections. I believe this is the situation  
in many of the listed multi-party countries.

In the US the president forms his own government (that is typically a  
single-party government) but in many other countries the government  
is not linked to the presidency (and may be a multi-party  
government). Therefore the presidential elections in many cases don't  
have much impact on the two-party vs. multi-party question.

The discussed two methods both favour large parties. Favouring large  
parties may lead to two party domination but not necessarily. In the  
case of the top-two runoff method there could be e.g. 3 or 4 parties  
that can make their way to the second round (depends on the country,  
its history, and the nature of the election). In countries where the  
president forms the government people are probably more loyal to  
their favourite big party than in coutries where the presidential  
post is more ceremonial or just "one man's post" (and the government  
will be elected via other routes). In the second type presidential  
elections the personal characteristics of the candidates play a  
bigger role (which leads to more parties having a chance).

In summary I'd say that IRV and top-two runoff favour big parties but  
having single-winner districts is a more important factor in making  
the two-party countries what they are. Condorcet, Approval and Range  
may elect centrist small party candidates quite easily but the two  
discussed methods tend to eliminate them. The differences between the  
two discussed methods are maybe not that radical. Combinations like  
single-winner districts + Condorcet would probably lead to compromise  
candidates in some districts but major parties (and centrists) would  
still have an advantage (when compared to fully proportional multi- 
winner methods).

Juho


On Feb 23, 2007, at 13:17 , Jan Kok wrote:

> The statistical evidence at http://rangevoting.org/TTRvIRVstats.html
> seems pretty good that IRV leads to two party domination in IRV
> elections, while (delayed) top two runoff tends to lead to a strong
> multiparty system.
>
> Why do those two methods, which seem strategically quite similar, lead
> to such different results? The above mentioned page has links to some
> speculations/explanations, which I find less than convincing. The main
> proposed reasons are:
>
> 1. Different strategy calculations by voters under the two systems.
> Voters who like a "third party" candidate seem more willing to vote
> for their favorite in the first round of TTR, than corresponding
> voters under IRV are willing to rank their favorite 1st. Why, why?!?
> Most IRV supporters in the US have no clue that voting their favorite
> 1st can ever hurt them. From my limited discussions with Australians,
> it seems most of them have no idea either. So why aren't Australians
> voting for third party candidates as their first choices, enough that
> they might occasionally win? While at the same time, voters in TTR
> countries feel free to vote for whoever they want, often enough that
> TTR countries tend to have flourishing multiparty systems?
>
> 2. Under IRV, if a "third party" candidate makes it to the last round,
> it draws little attention from the media. But under TTR, when there is
> a runoff, both candidates get equal attention for several weeks. This
> lets the finalists compete on an equal basis, so the third party has a
> realistic chance to win, and even if hse does not win, the party and
> candidate will be remembered in the next election.
>
> So, is either or both of those the real explanation(s)? Is there some
> evidence to back up these explanations? (I suppose statements by
> Australians, Irish, etc. saying that third party candidates rarely get
> any attention from the press, and statements by Argentinians, etc.
> that third party candidates _do_ get a lot of attention from the
> press, would be helpful.)
>
> It would be nice to put together some convincing and preferably
> non-technical explanation about why IRV leads to two-party domination
> and TTR does not. The purpose is to persuade IRV supporters to leave
> US TTR elections alone, or convert to something better than IRV.
>
> Thanks,
> - Jan
> ----
> election-methods mailing list - see http://electorama.com/em for  
> list info


		
___________________________________________________________ 
Try the all-new Yahoo! Mail. "The New Version is radically easier to use" – The Wall Street Journal 
http://uk.docs.yahoo.com/nowyoucan.html