[EM] IRV vs Plurality

[Jonathan Lundell]

On Jan 10, 2010, at 1:02 PM, Abd ul-Rahman Lomax wrote:

> At 01:06 AM 1/10/2010, Jonathan Lundell wrote:
>> On Jan 9, 2010, at 9:23 PM, Abd ul-Rahman Lomax wrote:
>> 
>> > On the other hand, in one-third of nonpartisan top-two runoff elections, which IRV supposedly simulates, the runner-up in the primary goes on to win the runoff, a "comeback election," according to a FairVote study. It simply does not happen with IRV.
>> 
>> It's hard to know what to make of this claim, other than, "so what?" -- if what you're comparing to TTR is the IRV candidates with the highest first-choice ballots.
> 
> I'm comparing *elections*. Yes, not all questions are addressed, but the response of "so what?" misses the point, obviously. If we assume sincere voting in TTR

I would not. TTR encourages favorite betrayal for much the same reason FPTP does.

> and IRV, and if we consider real world conditions (separate runoff, not merely some simulated runoff with the same voters and the same preferences and no new knowledge or shift of preference or better consideration, etc.), the results are quite different. That preference order doesn't shift in nonpartisan IRV elections, though the vote transfer process, was completely unanticipated, as far as I know. Mr. Lundell, were you aware of it? Did anyone predict this?
> 
> What this means that, with very little exception, you can use Plurality instead of IRV.

Again, I would not make that assumption. In this case, not simply because of the difference in voter behavior and voting rule, but because the nominating incentives are radically different.

> And this just may explain why Plurality has remained so popular for so long. In certain contexts, it works quite well.
> 
>> Presumably voters in TTR and IRV elections know the counting rules.
> 
> No. We can't presume that, in fact. Both TTR and IRV generally encourage sincere first-preference votes, not in reality, but in appearance. I believe that most votes in the current IRV elections in the U.S. are quite sincere, they aren't being gamed, as they would need to be to maximize voter expected satisfaction, in some cases. With TTR, indeed, voters may know the rules. They might even know IRV rules, but the *implications* of the rules, that knowledge, from what I've seen, is unusual.
> 
>> The meaning of a first-choice IRV vote is not the same as a TTR first-round vote. One plausible interpretation: if a TTR primary results in A>B, A tends to have all the votes A is ever going to get, but B, in the runoff, gets all the "anybody but A" votes, and wins.
> 
> When there is a runoff between two candidates, there is much closer examination of the two candidates. Further, preference strength enters into runoffs, causing results to shift toward what would be range voting results. Sincere range voting! I'm assuming, here, that the runoff is a special election. If the primary is the special election (seems to be common in North Carolina, for example), then it is the primary results that are range-shifted. That's improved election quality, Johathan. People vote when they care. We cannot tell from low turnout if the turnout is low because voters are equally disgusted with both candidates, equally pleased, or truly don't care about who wins no matter who the candidates are. High participation in a general election means a lot of noise, voters with low preference strength (which may be due to lack of knowledge, or may be intrinsic, i.e., they *do* know the candidates and don't care which one wins).
> 
> The proposed explanation for comeback elections, however, "anybody but A," is probably rare. Much more likely is this, as one example:
> 
> B is a dark horse, wasn't expected to do as well as actually happened. When B makes it into the runoff, suddenly those who supported B or who would have supported be had they known about B, will turn out to vote. In the other direction, those who voted in the primary and who actually have low preference between A and B may not turn out to vote; perhaps A was better known and got votes from these voters. So A may lose votes and B may gain them. Add to this the effect of additional campaigning, where B, now being a serious contender, gains more funding and more explicit support, perhaps the recommendation of other candidates.
> 
> Whether or not all this is likely may depend on the margin, but if B was truly a dark horse, B stands to gain votes from those who voted for other candidates. To me, the really interesting situation is where a center squeezed candidate doesn't make it into the runoff. What were the vote margins? Are there signs that the eliminated candidate was actually the Condorcet winner? By what margin and what preference strength? These would be factors in determining if a write-in campaign is appropriate. Using an advanced method in the runoff, even if there are only two candidates on the ballot, and allowing write-ins, allows fixing this problem with top two runoff! And the problem is less likely to arise if an advanced method is used in the primary as well.
> 
> 
>> OTOH, A>B>... in the first round of an IRV election tells us considerably less, since the cost of voting for a non-poll-leader is considerably less than with TTR. It's entirely plausible, though of course not necessary, that the A:B first choice breakdown reflects the lower-choice breakdown of the other voters.
> 
> This is what we see in IRV elections, when they are nonpartisan. It is as if the supporters of each candidate as most-preferred are a sample from the same population that is otherwise similar. So if, say, those who prefer someone other than A are considered as to their preference for B>C, the ratio of B>C preferences will be the same for those who prefer A. This was unexpected, to me, but it does make some kind of sense.
> 
> Most of our thinking about voting systems has been colored by thinking in terms of factional preferences, which are roughly predictable and not at all like this. We expect that a Progressive voter in Burlington will be more likely to support the Democrat over the Republican, and by significant margins, and we likewise expect that a Republican voter will be more likely to prefer the Democrat over the Progressive. In Burlington, the Democrats are in the middle, the center, they are the local moderates.
> 
> But nonpartisan elections don't seem to work like this. They are about personalities, less about issues and general stances on issues, I suspect. They may be about perception of trustworthiness.