My simplest 2016 anti-IRV pro-score summary

You are here

8 posts / 0 new
Last post
My simplest 2016 anti-IRV pro-score summary

Please feel free to reuse this text, I hereby make it CC0 public domain:


Rank choice FAILS to eliminate the spoiler effect, it just moves it. Consider that IRV claims it to be okay for Bernie to run as Green (as he was offered by Jill Stein) without spoiler. Well, let's say 45% top choice Trump, 35% top choice Bernie, 20% top choice Hillary. Hillary gets eliminated, because her voters are centrists, her 20% is split 6% 2nd choice Trump, 14% 2nd choice Bernie. Boom, Trump wins even if all the Bernie voters had Hillary as their 2nd choice. The claim that it was safe for all the Bernie voters to vote for Bernie because they would get their votes moved to their 2nd choice otherwise is a LIE. You only get your votes moved to your 2nd choice in IRV if you do not come in 2nd place overall. If your candidate comes in 2nd, your votes are worthless. Your votes can actually be WORSE than if you never voted at all in some cases…

As soon as people get this experience, they learn it's not safe to vote their real feelings, and 2-party domination is entrenched forever and we squander our chance at voting reform.

This ACTUALLY happened already in Bernie's own Burlington VT:

IRV is a failed system and so is any other type of forced-ranking.

Approval voting or the better full approach of Score (aka Range) voting has NONE of these problems, is basically optimal, and there's NO good reason to not have everyone get behind these good directions that *actually* eliminate spoiler effect. has all the details


(unrelated side note: this forum needs to remove most of the categories. It hurts the forum to have tons of inactive categories. Start with just a 3-5 categories at most and then add if things grow out of hand, the large number of categories is harmful to active discourse)

Needs a video

I really am thinking about making a video around this although my time is limited, it would take away from other timely and important projects. But we HAVE to respond to the moment here or we'll have everyone tragically pushing IRV!

Reviving an old thread

Reviving an old thread perhaps...

What are folks thoughts about FairVote's position? They claim that range/score/approval voting fails to meet the "later no harm" criterion and therefore voters cannot vote honestly and resort to strategic voting.
I was looking at this a few

I was looking at this a few months ago also. They also have this page:

The table there shows IRV having a high resistance to spoilers, and range voting a medium. But I don't understand how they arrived at that conclusion (the claim is unsupported on that page), and it seems backward to me.

They rehash their later-no-harm criteria, which also to me sounds specifically constructed to artificially favor IRV over range voting.

The objections presented here that I'd consider sensible are majority-rule and sensitivity to strategic voting. I personally disagree with their point about majority-rule, but I do think there is validity to the arguments relating to strategic voting.

The later-no-harm and various objections surrounding majority favorite seem to me to be multiple ways of stating the same basic idea, which is that the best candidate should be the favorite of the majority. They claim that "majority rule" is a basic tenet of democracy, and that this translates into majority favorite. But I think there are some hidden assumptions at play here. Another way to state this in a democracy is "the will of the people." I like this phrase better since it's less loaded with assumptions of how that will is to be calculated.

An economist would probably be more comfortable with range voting as an instrument of democracy because it maps directly to the concept of maximum utility. The range-voting winner is close to being that candidate that maximizes the social welfare function. I believe this is how an economist would define "the will of the people" and I find nothing objectionable about it.

Strategic voting: This can be a problem in two ways. One is when all voters are compelled to vote dishonestly for fear of being under-represented. The other is when one or more candidates can sway their own supporters into voting dishonestly in a way that gives them an advantage over other candidates. I fear that range voting is susceptible to both.

The first, in range voting, works like this. Strategic voters will identify leading candidates, let's call them the hopefuls, and vote so that their favorite hopeful is given a 10/10 and their least favorite a 1 (or 99 and 0 if you wish). The other "hopeless" candidates will probably get maximum and minimum scores, but could be scored in the mid-range, depending on how they relate to others on this stretched scale. I think that if all voters vote in this fashion, it's not a big problem. It's slightly worse for the population as a whole, but not much worse. This effect can be quantified with simulations under various assumptions. The problem here is that if only some portion of the population votes strategically like this, they are represented more strongly than those who vote honestly. That's not democratic since the tendency toward strategic voting is likely correlated with other political issues.

The second problem is a social mechanism. With range voting, a candidate does better for themselves by exaggerating the differences between them and their opponents in the minds of their own supporters. This is mud-slinging. We see this all the time and are mostly desensitized to the violence of it. But I would wish for an election system that did not reward this kind of behavior as much as range voting ought to. It creates unintentionally-strategic voters whose perceived utility functions are warped. This is going to be an issue with any voting system, but it seems to me like ranked systems may be a little less susceptible to it. and STAR voting

I've since become involved in the STAR voting movement around

Although it's an imperfect trade-off, the addition of a one-round automatic runoff to score voting helps address the arguments from FairVote. Regardless of their arguments being *bad*, it's simply a practical political matter to side-step the old debate.

There are some reasons to like STAR anyway. The equal vote argument is that everyone's vote has equal weight in the runoff.

Anyway, my updated simple point about IRV is:

IRV counts *some* voters' 2nd-choices and ignores the 2nd choices of others whose 1st choice still loses. All it's problems fall out of that fact.

STAR voting

Wow, I like that method very much! I've been intending to get back to some work on simulations, and would like to incorporate this and see if I can "observe" some of its benefits.

I had been thinking earlier about a related but poorer solution to the problem of strategic voting within a range voting system. I thought one could do a pre-election with the score cards, pick some top N candidates (those coming within some margin of victory), then go back and re-scale all the score cards making all voters equally "strategic." It seems like STAR voting accomplishes the nearly the same thing but in a less arbitrary and more palatable way. And the incentive to differentiate between leading candidates is clever! I like the proposed 0-5 scale, which should work well with that differentiation mechanic.

One benefit that range voting and some other methods have over both STAR and IRV is that simple vote tabulations can be carried over from local polling centers to perform the final tally and election. That allows for easier auditing (perhaps). But this is a technology issue and I'm sure good solutions can be found for it.

STAR works at precinct level partly

Tabulation of STAR can be done locally for the most part. Locally, you tabulate all the scores, just like in score voting. Also, once the total scores are added up from all the precincts, the precincts can themselves tabulate the runoff round too. So, the ballots and details *never* need to go outside the precinct level. Everything can be audited locally.

Log in or register to post comments

Follow The Center for Election Science on: