But for all the poetry, sometimes the prose makes a pretty good case, too. If there was ever an argument in favor of IRV's raw practicality, this is it:
For the US House race to fill the seat vacated by California's Rep. Ellen Tauscher (CA-10), Survey USA has released poll numbers gauging the strength of the 14 (yes, 14) candidates. Here are the results of the poll:
John Garamendi (D): 26 David Harmer (R): 18 Mark DeSaulnier (D): 15 Joan Buchanan (D): 12 Anthony Woods (D): 5 Chris Bunch (R): 4 David Peterson (R): 4 Mark Loos (R): 2 Other: 4 Undecided: 11Under this election's rules, the top Democratic, Republican, and independent finishers will go on to a runoff/general election to determine who finally takes the seat. The Swing State Project blog calls a Garamendi-Harmer race "a safe bet," and you can see why, as both candidates have strong leads over the other candidates in their parties.
But let's say this poll reflects the outcome on Election Day--is there any mystery at all as to who would win? The combined support for Democratic candidates in this poll is 58%, versus just 28% for Republicans.
In a situation like this, in which all the candidates from all parties are bunched up together in a single primary/preliminary election, it makes far more practical sense to use IRV--let voters rank their preferences, reach a majority consensus, and have the whole thing done in one single election, rather than hold a whole new, high-cost, low-turnout second election to ratify the obvious.
Oh, and all that stuff I mentioned about democratic ideals? That would still be there, too.