[EM] Does IRV elect "majority winners?"

[Abd ul-Rahman Lomax]

At 03:54 PM 1/2/2009, Terry Bouricius wrote:
>Abd,
>
>I think you miss-understood James Gilmour's question. He was asking about
>an exhaustive ballot election without any ranked-choice ballots. In his
>scenario 100 voters vote in the first round and 92 vote in the second
>round. Does the final round winner with 47 votes win with "a majority?"
>Robert's Rules and governmental election statutes would describe this
>candidate as a majority winner I believe.

Yes, if that was his meaning, that wasn't clear to me.

If the second round has *one* voter in it, and there is no quorum 
rule, and that voter casts a valid ballot, it's a majority, though it 
may be a constricted one. There are often voters in the first round 
who abstain in the second, and voters in the second round who did not 
vote in the first round.

Essentially, under Robert's Rules, the second round is a new 
election. The first election failed.

However, exhaustive ballot is not allowed under Robert's Rules, 
unless a contrary rule is specified in the bylaws. As to governmental 
public elections, exhaustive ballot is not used.

Runoff rules were not stated. If write-ins ware allowed in the second 
election, it is, in fact, *almost* in accordance with Robert's Rules, 
since no candidates are actually eliminated, only restricted from 
being printed on the ballot.

Does it matter *at all* how many voters voted in the *first round*? 
Suppose 10,000 voters voted in the first round, same candidate was 
eliminated. Then 47 voters voted in the second round, 92 voters 
total. The winner has (just barely) a majority.

Does IRV simulate exhaustive ballot? It's often been said, but a 
critical element is missing: IRV "denies" the voters the right to 
base their vote in the "runoff" on the results of the first election. 
Further, voters who may not have bothered to vote in the first 
election now may have increased incentive to vote, if their favorite 
is now perceived as having a chance of winning. Thus Le Pen gained 
about a million votes in the French 2000 Presidential runoff. But 
Chirac got all the rest of the votes....

Technically, top two runoff guarantees a majority winner if write-ins 
are prohibited. But I question whether this is a true majority 
winner. Rather, top two runoff does guarantee, legitimately, a 
majority winner *from the first ballot* or it fails and there must be 
a runoff. Where write-ins are allowed, a majority might not be 
reached in the runoff, it happens. Thus TTR doesn't truly guaranteeed 
a majority unless voters are coerced, prevented from voting with full 
sincerity, but only for, possibly, a lesser evil. It's a compromise.

What is true is that TTR *seeks* a majority, and that it usually 
finds it. Where there is Condorcet failure (due to Center Squeeze in 
the primary, basically), though, it fails, in *substance*, and so may 
IRV under similar conditions. However, a better primary method, and a 
better runoff method, would find a majority far more often, only 
rarely failing.

IRV *pretends* to find a majority. Real runoffs do find a majority 
(if they are limited to two candidates), but from a constricted context.

What FairVote advocates have done is to try to conflate the two 
elections. Then they imagine that those who didn't rank additional 
candidates have "abstained," trying to connect this with those who 
don't vote in a real runoff.

There are a number of false assumptions here. One is that the voter 
set is the same in TTR and in IRV. It isn't. The other is that voters 
were even able to add additional preferences; *in theory*, if full 
ranking is allowed, then it does become *possible* to argue that 
voters "voluntarily abstained." But full ranking isn't allowed in any 
but the smallest IRV elections in the U.S. Usually it's three ranks, 
with as many as twenty candidates or more.

Top Two Runoff allows voters, where a majority isn't plain from first 
preference -- which it usually is -- a closer look; different voters 
are motivated to vote, and the relative vote in the two elections is 
irrelevant.

Sometimes it increases. My observation is that this happens most 
often when the primary makes a poor choice, due to Condorcet failure. 
Voters, then, seeing an election between mediocre and total disaster, 
turn out in increased numbers. Otherwise, ho hum, it's an election, 
many will think, between Tweedledum and Tweedledee, and it just 
doesn't matter to them, so they don't vote. But increased turnout can 
also happen when a dark horse makes it into the runoff and attracts a 
lot of additional support.