Instant runoff voting method
I honestly think that we do IRV wrong. It’s frequently used as part of a first-past-the-post system (FPTP), and that in itself is known for causing pendulum swings with growing levels of extreme policies, which end up jerking the population between ever deeper extremes.
This is something I have been saying for years. I first thought of it when looking for solutions to the issues people had raised with the Bucklin voting system.
Now, forget that term/voting method. This issue applies to most, if not all, IRV uses today.
I think of this again today as I look at how divisive the United States is today. I come to you today asserting that if you want to “Make America Great Again,” We need a unified, not a divided, front.
In theory, the IRV systems allow the most wanted, and hopefully most qualified candidate to win.
Saying no, “people are dumb,” would also invalidate any vote you would have.
Ignoring people that are “dumb” is no different than ignoring people that are “not dumb.” Voting is the way people are heard. People not heard lead to revolutions.
The pro of IRV
Anyway, here’s the deal. With the use of IRV, we can feel more comfortable having 3rd candidates. This eliminates the pendulum swings of FPTP voting systems. (See FPTP issues)
Gaining a third candidate introduces new ideas and movements into the system, allowing us to have the option for a candidate who can fix things. Perhaps they’re more centrist, but that unifying factor is what can help provide a unified front versus a divided one.
Don’t worry, the views on either end of the spectrum still get heard. Remember, they still have to engage you and your opinions. Additionally, there are more players in the game, including people from other districts and levels of government.
After all, we have three branches of the federal government. All pick the POTUS; Congress members come from local districts, and Senators generally represent your state; and those groups pick the Judges for federal courts. NOTE: There may also be issues to address with the Senate & House (See: Congress: House & Senate Improvements).
Additionally, at the state level, we have the same branches locally: Legislative, Executive, Judicial branches.
Finally, locally, you have City/Town and County governments that can address needs. All three levels of government work together, often challenging each other, while working together to do the best they can for the people they govern.
The con of FPTP/First-Past-The-Post voting
Basically, when a third candidate enters the FPTP system, people are afraid to vote for them fearing they are throwing their vote away.
The concept is:
Part 1: Let’s assume each candidate will receive 33% of the vote. Now, if my candidate (Candidate C) were to potentially receive only 33% of the vote, because they are not part of one of the major parties, they wouldn’t have the support to overcome the other two, so that 33% ends up being 20%. Now, the other two candidates get 39/42% of the vote. (Candidates A/B)
Part 2: Now we have voters strategically voting. lets call it shooting the line or strait line voting. Of that 20%, ten percent opt to vote for the candidate they think “CAN” Win. Perhaps it’s splitting 4% & 6% making the other two candidates have 43% & 48% (Candidates A/B) while…
Part 3: The final 10% stick to the “BEST Candidate” even though they won’t win. Maybe it’s to make a statement, maybe to be loyal.
What statements are they making?
A: Is it that they like their views and want to make sure the winners consider them?
B: Is it that neither of the other candidates is decent enough to deserve a vote?
Is that right? Here’s the outcome with IRV
Finally, here’s the deal! Of that last 10%, nine percent would prefer Candidate A. Thus, in theory, that 43/47% is actually 52% vs 48%. Analyzing these results, Candidate A is the preferred winner, yet Candidate B is the chosen winner, winning with ONLY 47% of the vote.
Analyzing the issue with IRV
With IRV, one would say the solution is to let people vote for a second choice, and allow their votes to roll over.
Looking at it, because they’re not throwing their vote away, the voters are not afraid of throwing away their vote, so the vote sticks to 39/41/20%. Then IRV kicks in, eliminating Candidate C and redistributing the second-choice votes from their voters to Candidate A or B, respectively. Thus, Candidate A wins with 52% of the vote.
YAY, The best candidate wins!!!
However, did they really? If you analyze the votes again, we see that 50% of those who voted for Candidates A & B also liked Candidate C; however, because Candidate C was not a part of a major party, they discounted their viability. Hence, the drop from 33% to 20%
Let’s further analyze this:
Method 1: Round 1>> Final Outcome >> AV totals
Candidate A: 33% >> 16% >> 52% (Loser in FPTP)
Candidate B: 34% >> 17% >> 48% (Eliminated via AV) (FPTP Winner)
Candidate C: 33% >> 65% >> 65% (Winner via AV-IRV) (Eliminated via IRV)
Method 2: Round 1>> Final Outcome >> AV totals
Candidate A: 39% >> 19% >> 52% (Loser in FPTP)
Candidate B: 41% >> 20% >> 47% (Eliminated via AV) (FPTP Winner)
Candidate C: 20% >> 59% >> 59% (Winner via AV-IRV) (Eliminated via IRV)
^^^ These are wrong…. See “Why AC” below, where I return Candidate A & C’s votes when B is eliminated
Either way, Candidate C was approved or even liked by at worst 59% of the voters and possibly even 65% of the voters. Yet, we eliminated that choice up front!!!
Candidate C was eliminated with both IRV & FPTP, despite being the most liked Candidate by a margin of 7-13% in this example.
The Solution IRV via Approval Voting (AV)
NOTE: I’m not a fan of the strict AV voting method. Its too easy to result in a tie as eventually, everyone approves of several people.
Instead, use AV to eliminate the least approved candidate, and then redistribute the votes just as you would in a traditional IRV scenario.
Remember how I mentioned straight-line voting. The concept is that you vote only for the ONE candidate you like, so that you don’t dilute your vote, potentially allowing another person to win, while you really only want Candidate X to win.
Nice in concept, but that’s the same issue as throwing away your vote that happens when you vote for a third-party candidate in the original FPTP system.
People have attempted to solve the issue of throwing away your vote with IRV by “REQUIRING” People to make a second choice. This was ruled unconstitutional, hence the issue with Bucklin voting. People found it worked best if they always cast a second-choice vote, and rather than reworking it, they just tossed it away.
Yet, the issue still stands. IRV often eliminates the preferred candidate because people shoot the line to eliminate the possibility of their candidate being the 3rd candidate in the first round. i.e. to prevent throwing away their vote.
Why AV?
I came up with this because of the issues with forcing people to make a second choice. By using AV and changing the elimination methods, I believe this encourages people to make that second choice. After all, if their second choice means that Candidate B instead of C is eliminated, maybe their choice of A will still win.
Check this out. Let’s look at methods A & B again, but split the choices.
In these Examples, it doesn’t mean the 3rd party candidate wins. it just means the least popular is eliminated vs just the lowest vote getter in round 1
- Candidate A’s & C’s 2nd votes are returned when Candidate B is eliminated
- Thus, a new race just between Candidate A & C
- Tie Breaker, if needed, is AV? The constitutionality of AV for the outcome or even the elimination round is in question due to the one-person-one-vote concept.
Playing with numbers to show that Candidate C might not win even if surviving round 1
Remember: 2nd votes from voters of candidates not eliminated don’t count for other candidates in the final count since their candidate isn’t eliminated.
| Example 1 Winner: Candidate | (Equal due to IRV, but also Dead heat all the way) — Rounding in favor of Candidate C — 51% of B’s voters like A (21/20) — C voters split on Candidate A & B | |||
| Example 1: | Round 1 | AV Totals | Final Outcome | |
| Candidate A: | 33% | 16c+17b+33a=66% 16c+16b+33a=65% | 0c+0b+ 0a=0% IRV: 16c+0b+33a=49% | (Loser in FPTP) (Eliminated via AV) (Loser in IRV) |
| Candidate B: | 35% | 16c+35b+16a=67% 16c+35b+15a=66% | 0c+35b+16a=51% 0c+35b+15a=50% | (FPTP Winner) (IRV Winner) |
| Candidate C: | 32% | 32c+18b+17a=67% 32c+19b+18a=69% | 32c+0b+17a=49% 32c+0b+18a=50% | (Eliminated via IRV) |
| In a close race, is this fair? Candidate B got 3% more in Round 1 Candidate C got 3% more approval | ||||
| Where’s the tie breaker? — Just as NOW, Tally their votes, then just as with IRV, tally the 2nd votes of the eliminated candidate. There’s bound to be a difference, just like today has close races. Just like now, a true tiebreaker is done via the sitting VP? Thus, a trend towards the incumbent/sitting party. | ||||
| Example 2 Winner: Candidate A | (Voting for preferred candidate up front) — All of A voters like C — 51% of B’s voters like A (21/20) — C voters split on Candidate A & B | |||
| Example 2: | Round 1 | AV Totals | Final Outcome | |
| Candidate A: | 39% | 10c+21b+39a=70% | 21b+ 39a=60% | (Loser in FPTP) (Winner in IRV) (Winner via AV-IRV) |
| Candidate B: | 41% | 10c+41b+0a=51% | 0+0+0=0% | (FPTP Winner) (Eliminated via AV) |
| Candidate C: | 20% | 20c+20b+39a= 79% | 20c+20b=40% | (Eliminated via IRV) |
| While Candidate C has the highest AV percentage, Candidate A is higher in round 1, and more of Candidate B voters prefer Candidate A. The Winner goes to Candidate A: 60/40 | ||||
| Candidate C still loses, but Candidate B only had 51% of the vote at any time. It’s much preferred to take their 21% vote to A and 20% to C, both of which are more popular in general | ||||
| Example 3 Winner: Candidate C | (Voting for preferred candidate up front) — All of A likes C — 75% of B’s voters like C — C voters split on A & B | |||
| Example 3: | Round 1 | AV Totals | Final Outcome | |
| Candidate A: | 39% | 10c+10b+39a = 59% | 10b+39a =49 | (Loser in FPTP) (Winner in IRV) |
| Candidate B: | 41% | 10c+41b+0a = 51% | 0+0+0=0% | (FPTP Winner) (Eliminated via AV) |
| Candidate C: | 20% | 20c+31b+39a = 90% | 31b+20c=51% | (Eliminated via IRV) (Winner via AV-IRV) |
| Candidate C is very popular with 90% approval, and would have been eliminated in both FPTP & IRV methods of voting. Granted, they still barely won, but Candidate 2 was still liked by many of Candidate B’s voters; thus, their votes were not wasted. If the vote counts had been slightly different, so that Candidate A had won, it would have been justified, even with Candidate C having a 90% approval rating. Especially given that Candidate C was at least given a fighting chance that would not have been available with FPTP & Traditional IRV. | ||||
| This was an exhilarating and competitive race. However, with the Lowest Approval rating, Candidate B’s winning in FPTP & IRV would have been an injustice to the majority of voters. | ||||
| Example 4 Winner: Candidate B | (Voting for preferred candidate up front) — A Voters split between B & C — 75% of B’s voters like C — C voters split on A & B | |||
| Example 4: | Round 1 | AV Totals | Final Outcome | |
| Candidate A: | 39% | 10c+10b+39a = 59% | 0+0+0=0% | (Eliminated via AV) |
| Candidate B: | 41% | 10c+41b+19a = 70% | 0c+41b+19a = 60% | (FPTP Winner) (Winner via AV-IRV) |
| Candidate C: | 20% | 20c+31b+19a = 70% | 20c+19a=39% | (Eliminated via IRV) |
| Candidates B & C are both very popular with 70% Approval. C would have been eliminated with 20% of the vote in either FPTP or IRV, but their votes counted instead, and Candidate A’s votes were not wasted as they still counted towards Candidate B & C | ||||