[EM] a comment

[Abd ul-Rahman Lomax]

At 12:20 PM 4/20/2013, David L Wetzell wrote:
>If you're going to pit two election rules against each other by 
>using them both and then have voters decide between the cases when 
>they differ then you're going to have sample selection problems.

The "comment" seemed to assume public elections. Voting systems can 
be tried in NGOs, and that's where the future lies, my opinion. It's 
very unlikely that we will see major voting reforms take place in 
governmental election systems without them having seen usage in NGOs.

Having said that, history isn't necessarily friendly to my idea.

Bucklin voting was all the rage in the period 1910-1920 and a little 
later. Yet I never heard of it being used outside of public elections.

It worked in public elections, no pathologies were asserted at the 
time other than that it allowed a runner-up in the first preference 
votes to win the election. That was considered horrifying to the 
Minnesota Supreme Court, which, effectively, interpreted the state 
constitution as *demanding* plurality. Very strange.... (FairVote 
later argued differently, but I'm quite sure they would have 
disallowed IRV just the same.)

The only problem was that in nonpartisan elections -- party 
primaries, much later -- it frequently failed to find a majority at 
all. That wasn't Bucklin's fault; IRV would have failed even more. 
The real fix to that problem would have been a runoff, and what was 
*actually done* was to dump Bucklin and to use top-two, vote-for-one 
in the primary, with a runoff when no majority was found. If they had 
simply used a hybrid system, say a Bucklin primary, with a runoff 
when needed, history might be different.

But Bucklin had been sold the same as IRV more recently: find a 
majority without expensive runoffs....