Of course that is best addressed by having an election system with good integrity properties which presumably you will get around to proposing here.

—-

The current wikipedia definition of IRV is actually a bit misleading in that it abuses the term majority.

>”This process is repeated until one candidate has a majority
>among votes for candidates not eliminated.”
It should read ‘a majority of remaining first choice preferences’.

1. Other preferences that voters expressed are ignored. So IRV will tend to not elect compromise candidates who have broad support.

2. Individual ballots can be exhausted. This is especially likely when truncated IRV is used rather than full IRV. Exhausted ballots are no longer counted when deciding who has a ‘majority’.

But since you interest is in election tabulation and integrity, I’ll focus on the administration issues.

What is IRV?

From wikipedia:

Instant-runoff voting (IRV) is a voting system used for single-winner elections in which voters rank candidates in order of preference using one ballot. No further voting rounds are required. If no candidate receives a majority of first choices, the candidate with the fewest number of votes is eliminated, and the votes cast for that candidate are redistributed to the remaining candidates according to the voters’ indicated preference. This process is repeated until one candidate has a majority among votes for candidates not eliminated.

—

The nature of IRV results in complicated administration. This is because of two factors:

1. It collects a lot of information from each voter. The system described above collects 3 ranks of data. Votes that mark the same candidate in multiple ranks or have multiple votes at the same rank or that skip a rank might all be considered spoiled depending on the election rules. IRV is often paired with election machines since machines can reject spoiled ballots quickly enough for the voter to correct their vote. This same issue also applies to other ranked election methods such as Condorcet methods but is not as bad when the election method allows ties at a particular rank.

2. Results are not summable in precincts. This is fairly unique as almost all single winner election methods. To figure out who the winner is you either need to gather all the ballots in a central location or to have a multi pass system where precincts report initial counts to a central location and then once all the precincts have reported this the central location directs all the precincts to eliminate a candidate and adjust their counts. Rinse and repeat.

http://minguo.info/election_methods/evaluation/sc_summability_criterion

IRV is an awful single winner election method.

I vastly prefer approval voting.