Unless we enact some form of ranked-choice voting, we will never be able to break out of the political Republicrat trap, and never have a chance at getting the real “change” that people desperately want. Third-party candidates will continue to be marginalized by the media, who are increasingly concerned with picking (or making) winners rather than conveying information. People will continue to be afraid to “waste” their vote on a candidate who seems unlikely to win — as if voting is about predicting the winning horse rather than expressing what you want from government. And they will continue to hold their nose and vote for the “lesser of two evils,” lest they tip the election to the candidate they like least.
It’s almost impossible for real outsiders — other than show-business celebrities and self-funded billionaires — to get a foothold.
There are several forms of ranked-choice voting, with various pros and cons. But it would be better than the current situation, where pols don’t have to worry about being held accountable because they know people really have no choice but to ricochet back and forth between the two big-party options, whose offerings, in practice, are all too similar, and brought to you by the same corporate sponsors.
Ranked choice voting legislation languishes in the State House, because the insiders want to keep the game to themselves. We’ll need a ballot initiative to break it open.
You might have finally hit on an idea that progressives can get behind for public vote. I generally don’t like legislation by ballot box, but this is fundamental enough to how our system of government works that I would suggest proposing it as a constitutional amendment. I do believe that constitutional amendments should be enacted by the people as the ultimate sovereign.
I would also like to see some way to vote for more than one candidate and perhaps express preference, but I really don’t like instant runoff voting (IRV) and am disappointed that was the method chosen by the people who were pushing a ballot initiative earlier this year. IRV is complicated to implement in large elections because you have to have access to all ballots at counting time and have to make sure that all voting machines have a way to express the preference. It also has the potential to create nightmarish recount scenarios in a very close election.
<
p>The simplest thing to do would be to simply let people vote for as many candidates as they want to. That should be very easy to implement and easy to count. It would also eliminate “multiple votes” as a reason to invalidate a ballot during recounts. Under such a system, I think we would have seen different results in many-way races such as the recent 10th Suffolk State Rep primary.
but let’s say you’ve got four candidates:
<
p>Alice Awesome
Patty Pretty-good
Dee Dee DePlus
Wilma Worst-candidate-ever
<
p>
<
p>Now, who do you vote for? A for sure and definately not W. But what about P and D? Well, it depends.
<
p>If it looks like a horserace between A and D, you might vote for A and P only. If, however, W is doing well, you might vote for A, P, and D.
<
p>What if A and P, and W are polling at 30% and D is polling at 10%. You’ll vote for A for sure and not for W, but what of P? On the one hand, you’d rather A win instead of P, and they might run close. On the other hand, you don’t want W to win, so you’ve got to vote for P too.
<
p>
<
p>It seems to me that voting for all candidates you like doesn’t really eliminate the difficulty of tactical votes, at least not in all cases. I do agree it would be better. For my money though, IRV still works better. You get to differentiate between candidates, which means you provide more information to the body politic. Furthermore, the concern about access to all voting machines isn’t really that hard. Each machine scans the votes and knows how many ballots are ABC, ACB, BAC, BCA, CAB, and CBA (for example). It’s simply a matter of counting how many of each pattern and sending those counts “to the mothership”. This way anyone could reconstruct the runoff portions easily enough. Certainly manageable, and no harder to calculate than situations now with multiple wards/precincts.
Yes, you can always construct a scenario where some given system does better, but I think that simply allowing multiple votes will solve most of the problems we are really trying to address.
<
p>If you want to give preference, you can simply let people vote for each candidate multiple times, which is still a much simpler system to understand and operate than IRV.
My favorite ice cream flavor is chocolate.
My second favorite is chocolate chip.
My third favorite is mint chocolate.
My fourth favorite is chocolate chip cookie dough.
<
p>
<
p>She can do this all day.
She understands ranking her own personal choices, she definitely does not understand the IRV voting procedure.
IRV has been proven an effective reform in jurisdictions large (Australia) and small (San Francisco). The implementation issues you point to are easily surmountable. Voting machine manufacturers are increasingly deploying machines capable of capturing ranked preferences, due to the increasingly demand, and also know how to cheaply retrofit some existing machines, including the Optech Eagle IIIP, used throughout much of the state. Also, IRV does not require centralized tabulation (what I believe you meant by “access to all the ballots at counting time) — a common erroneous claim made of large scale IRV elections. A statewide recount, should one be necessary, wouldn’t be noticeably more nightmarish with IRV than with the current system, and is a pretty weak argument in face of the most important problem with the current system: it frequently elects the wrong person.
<
p>The system you are suggesting is more commonly known as “Approval Voting.” It is easier to implement at the machine level, but moves all the complexity into the hands of the voter. What is the criteria one should use to determine whether to put a vote next to a person? This is a non-trivial question, some of the expert answers to which involve complex mathematical formulas, formulas which require access to highly accurate polling data to access the probability of a candidate being elected. Stomv highlights some of the mental gymnastics one has to do to vote correctly under Approval. Here’s another example: consider an election between Barack Obama, Hillary Clinton, and John McCain, and consider the dilemma faced by a Democrat that can vote for as many people as s/he wants, all votes equally weighted. The decision to bullet vote for your favorite Democrat or vote for both Democrats is agonizing and given enough Democratic bullet votes, McCain can win even if Democratic voters are a majority of the electorate.
<
p>In IRV, the strategy is simple: rank your candidates in order of preference. If 31% rank Obama>Clinton>McCain, 29% rank Clinton>Obama>McCain, and 40% rank McCain first, then Obama wins. With Approval, if enough Obama and Clinton supporters bullet vote, they could throw the election to McCain. This strategic problem with Approval Voting is known in the voting systems literature as the “Burr dilemma.”
The more complicated the voting process, the more you depend on computers and the harder it is to verify that the computers implement process correctly. The simpler the process the better. I believe a more complicated system is only justified if it can definitively be shown to solve an important problem that is not addressed by a simpler system. IRV does not get over that hurdle for me.
<
p>Let’s put a simpler system in place, and see if that is enough, and only then consider whether something like IRV is appropriate.
<
p>
and count each ballot by hand. Polls close 10pm on a Thursday, counting begins 9am Friday, and finishes at about 2am that night.
<
p>IRV doesn’t need computers, they just make it faster.
I don’t care what other country is doing it, as I have said, I believe IRV is more complicated than is necessary to address the problem.
<
p>People who dogmatically push IRV are making the perfect the enemy of the good.
IRV can solve a very real problem that is not solved by your suggested system: the spoiler problem. Nearly everywhere Approval has been tried, it has been abandoned for the very reasons I outlined. It is known to not work.
<
p>Second, I dispute your claim that it is “simpler.” It moves the complexity from the ballot into agonizing strategic decisions for the voter in the booth.
<
p>Lasty, and perhaps, most importantly, IRV in practice is actually more thoroughly auditable and verifiable than either Approval or our current system. The reason is that in practice, verifying an IRV election demands re-counting from the ballot level, rather than simply checking the recorded precinct sums against the machine. Read audit instructions here. It is no surprise that the first IRV elections in San Francisco (CA), Burlington (VT) and Aspen (CO) were the most transparent, verifiable public elections ever held up to that point. Nor is it a surprise that the founder of Ireland’s election integrity movement is also a big fan of ranked voting.
But approval voting is used by the Mathematical Association of America and the American Statistical Association. And I trust the MAA and ASA to be better at math and statistics (and a voting system is just math and statistics) than Dartmouth.
<
p>You also mentioned three places that have used IRV; Burlington, Aspen, and San Francisco.
<
p>But you failed to point out that voters in Burlington and Aspen have both recently voted to get rid of IRV. I’d be more worried about a municipality than an alumni association.
<
p>But appeals to popularity shouldn’t be what concerns us. I prefer to appeal to data.
<
p>http://rangevoting.org/BayRegs…
<
p>And the data shows that approval is significantly better than IRV.
First, your example is entirely flawed and shows a deep misunderstanding of how approval voting works.
<
p>Voting for a second candidate–the other Democrat in your example–can never CAUSE the one Republican to win under approval voting. If the Republican were already going to win, your vote could cause your second-choice Democratic candidate to win instead (and not voting for your second-choice Democrat could leave the win with the Republican), but if the Republican were not already winning, it is IMPOSSIBLE for you to make them win by voting solely for Democrats.
<
p>Sure, it might be hard to decide whether or not to approve your second choice (the Burr Dilemma is real), but you have completely mis-characterized the risks. The risk is either: by not voting for your second choice, your last choice wins, or: by voting for your second choice, they win instead of your first choice.
<
p>Comparatively, the risk under “instant runoff” is the same as the risks under the current plurality method: by ranking your true favorite first, your least-favorite wins.
<
p>Would you rather fret over 1st and 2nd, or over 1st and LAST?
<
p>Granted, for this to be a concern, there have to be three strong candidates in the race; but if there are, the “simple strategy” you posit for IRV, fails; just like under plurality. An example:
<
p>45%: A > B > C
10%: B > A > C
15%: B > C > A
30%: C > B > A
<
p>In this example, by IRV, A wins. But, if a few of the C-first voters were to vote tactically, by putting B first instead of C, like this:
<
p>45%: A > B > C
10%: B > A > C
20%: B > C > A
25%: C > B > A
<
p>Then the winner changes to B. In other words, by voting for the lesser of two evils, they avoided having the greatest evil win. This is exactly how voters have to vote today. Which is why Australia has, just like the US, two dominant political parties, despite using IRV. This is why San Francisco saw the same party-makeup on its city council both before and after changing to IRV: IRV doesn’t change anything.
<
p>Under approval, it is always in the voters best interest to approve of their honest favorite, and to not approve of their honest least-favorite.
<
p>But under IRV your best interest are NOT always served by ranking your honest favorite first.
<
p>And you’re also wrong about needing access to all ballots to adjudicate the election: IRV does need that. When you break an IRV vote into precincts, it’s possible that one candidate wins in every precinct, but when you add all the precincts together a different candidate wins.
<
p>IRV is a useless, false, non-change. It doesn’t work. Approval voting isn’t perfect (nothing is) but it is far superior to IRV; a fact demonstrated by extensive computer modeling: http://rangevoting.org/BayRegs…
Hi Shirley,
<
p>Good to see you active. I have been active too in my own KingCastic way.
<
p>Enjoy the video from yesterday’s rally:
http://www.youtube.com/watch?v…
http://seminal.firedoglake.com…
http://christopher-king.blogsp…
“There will be no Arizona in Massachusetts!” echoed hundreds of revelers of all races, religious, ethnic and national backgrounds at the South End’s Titus Sparrow Park, the first major campaign rally for incumbent Massachusetts Governor Deval Patrick and Lieutenant Governor Tim Murray.
<
p>To the folks I met today, the background on the heinous Immigration policies of New Hampshire vis a vis the nefarious relationship between Arizona Sheriff Joe Arpaio and NH Senate Candidate Kelly Ayotte Ayotte is here.
<
p>Be sure to follow the excellent work of Arizona New Times’ writer Stephen Lemons right here, we clearly have mutual respect for each other and for Arizona’s Dennis Gilman, here is Mr. Lemon’s comment on Gilman and Yours Truly.
<
p>The ongoing CORI issue seen in today’s fliers that His Excellency is working to clear up is here.
*****************
Meanwhile, Boston area Pop Folk and Peace Legend James Taylor had a nice soul shake for me (I’ve never had the pleasure of meeting him in person though I’ve been a a show of his in the 90’s) and some stirring reminders those who would like to blame Governor Patrick for all the World’s woes:
<
p>
<
p>****************
<
p>The video is a wake up call Moderates who are considering a GOP/Tea Party vote and to Dems who mamby-pamby the evil that the GOP represents.