Someone mentioned how Maine is having a ballot initiative to pass instant runoff voting.
How can we get that kind of system in the US, that applies on all levels of government (local, state, federal) elections.
Aside from ballot initiatives, is there anything that can be done? Neither of the 2 main parties is going to support runoff voting. So getting it passed in congress will not be realistic.
What about the courts? Could an argument about runoff voting being constitutional be made to the courts to change the electoral system?
Ballot initiatives
Legislative changes (highly unlikely)
the Courts
The only way to get in every state at every level would be either a federal constitutional amendment or legislation of some type at the state level in every single state (which might require a constitutional amendment and referendum in some states). Not all states have ballot initiatives, so it would require the legislature to act in some.
Congress could require it for elections to the House and Senate and maybe for presidential electors, but not state and local elections.
I guess you could try to get the Supreme Court to strike down the electoral system used in the vast majority of elections nationwide. Good luck with that. My guess is that the backlash would be in the form of a 28th Amendment.
You can’t because it’s not in the interest of either party to lose their duopoly on the political process. Violent revolution then re-writing a new constitution is the most realistic way it could ever happen (and by realistic I mean incredibly unlikely but still more likely than the Dem’s and GOP agreeing to change to instant run off voting).
IRV doesn’t really threaten their duopoly. It actually protects the major parties from competition with minor parties. Right now a party has to keep tabs on what its fringe is thinking and try to keep them from voting for a spoiler. With IRV they can ignore them and just take their preferences.
That’s really not how it works in practise in Australia where we have instant run off voting. One of our major parties is actually a coalition (The Liberals and the National Party, oh and our liberal party is actually conservative The Labour party then quite often has to cooperate with the Greens or the Democrats in certain areas.
Most likely if the US did have instant run off voting the Democrats would immediately split into the far left and a clinton centrist party, they’d give preferences to each other over other parties but compete on the ballot. Similarly the GOP would split into the sensible business right and the loony right, again they’d give preferences to each other but otherwise compete. You’d end up with two loose coalitions with no one party able to get enough to win on it’s own. Now I think that would actually be a good thing, but it would be the end of the power duopoly that they now have.
The first major party to fragment will cede primacy until the schism was healed.
Or even the meeting of political minds occurring between the two central factions.
You have this curious view that the world would be better if it was carved into four even segments. I don’t see any measure why that would the case and your envisaged splits would more likely leave two dominant central factions and two outer wing and increasing disaffected rumps.
Why would any political party fracture just so it could feed votes from one faction back to the other? The centralists spending cash and diluting their message so as to fight two opposing ideologies simultaneously? The outer parties spending their cash and political capital cannibalising their nearest ideological neighbour. The very definition of political lose-lose.
Actually, there is a way it could happen short of that.
The Australian experience with IRV has been that two major parties dominate Parliament anyway. If voters revolt against nominees like Donald Trump in large enough numbers that the GOP can’t keep winning states reliably, then it’s to the benefit of the GOP to adjust the system so that the dirty rotten country-clubber they nominated doesn’t have to be the people’s favorite, just preferable to the untrustworthy Muslim-coddling baby-killer on the Democratic side. IRV, within what is otherwise the present system. would help GOP control survive. And since the GOP still control Congress, they can do it.
Of course, they won’t, first because it’s “new & foreign,” not “traditional American,” and second because it’s too hard to explain to the legislators who’d have to vote it in. (Note that I am not saying that ranked voting is a hard concept. I am saying that legislators are not chosen for their understanding of obscure election science and exotic maths. They’re chosen to be “spokesmodels” more or less.) They’re just going to have to rely on traditional methods like propaganda, gerrymandering, ballot-box stuffing, and voter intimidation.
I’m not sure why instant runoff voting is put up for consideration instead of approval voting.
Approval voting requires no changes in the ballot (each candidate is either voted for or not, just like now), is simply to explain (vote for as many or as few candidates as you wish), and is easy to evaluate the results (whoever gets the most votes wins).
Instant runoff voting requires ballot changes (each candidate box may get a number), is more complicated (you have to rank the candidates), and is tricky to evaluate (sometimes a candidate will do better if you rank them lower). And IRV magnifies recounts: near ties at any point in the runoffs requires a recount.
Of course proposing a more complicated voting system might be part of the strategy of those who don’t want to change the system.
Congress can set the conditions for electing Senators and Representatives (Article I, Section 4), but not for Presidential electors (Article II, Section 1: “Each state shall appoint, in such manner as the Legislature thereof may direct, a number of electors…” - notice that it says nothing about requiring a vote of its population, but that’s another story).
What effect does IRV have in the House in Australia? I expect it to have more of an effect in the Senate, where six Senators at a time are elected from the same ballot in each state. (In the USA, in the rare instances where both of a state’s Senate seats are up for election, the two elections are handled separately.)
My theory of the moment: approval voting would be treated as first-past-the-post by most people, since a vote for somebody besides “your” candidate makes it more likely that your preferred candidate loses.
Also, I have seen petitions calling for an election system where, in a “vote for up to X candidates” election (e.g. city council, school board), you can cast multiple votes for the same person. I have a feeling that approval voting would turn into something like this.
IRV is better than approval voting because it gives more information, and the more information you feed into a system, the easier it is for the system to produce the best outcome. I would like to be able to say, for instance, that I prefer Sanders over Clinton, either of them over Kasich, Kasich over Jeb, and Jeb over Trump. If I’m using approval voting, where should I draw my approve-disapprove line? The most logical place I can think of is in the biggest gap between candidates, but that means just drawing the line between the two parties, and I can’t see any way that produces better outcomes than our current primaries-then-general system. With IRV, though, I can say that I prefer all Democrats over all Republicans, while still also conveying which Democrats and which Republicans I’d prefer.
That said, though, there are other voting systems which are even better than IRV, in a mathematical sense that their paradoces of voting are a strict subset of the paradoces found in IRV. The main problem with them is that they’re more complicated to explain, and a voting system is no good if it doesn’t have the trust (which requires understanding) of the people using it.
Burlington, Vermont, Bernie Sanders old stomping grounds, implemented IRV several years ago and then repealed it a few years later. It is not the panacea its advocates think it is, and it can easily lead to perverse outcomes.
Although there is no perfect voting system, I would rather see non-instant runoffs as occur in France and Louisiana.
The effect of preference voting in the Australian House of Representative is to bolster the major parties.
The number of independent/third party candidates who win seats is small, rarely more than 5 from 150. However once established a independent can be hard to dislodge.
The multiple member proportional voting system in the Australian Senate makes it virtually inevitable that minor parties have disproportionate (over) representation and the balance of power.
To quote ex PM Paul Keating “unrepresentative swill”
That’s why I like “graded” voting. Almost everybody in America has experienced, or is at least familiar with, letter grades, A B C D F, as in school. Assign any grade you like, independently for each candidate in the race: five-point range voting.
Why wouldn’t voters simply vote at the extremes in any range scheme? If my real opinion is that candidate X is a C and candidate Y a D, I’m going to vote X=A and Y=F just to maximize my voting power. It seems equivalent to approval voting in this respect.
I was a big fan of IRV until I was looking at the recent Australian results. Two big parties with the balance of power in cross-benchers? Meh.
I’m more interested in party-list systems now. Greece has an interesting one. Of course, Greece also ran budget deficits for a few decades in a row, but that’s a problem the USA already has.
This is a common misconception we at the Center for Election Science call the “expressiveness fallacy”.
In short, the IRV tabulation algorithm simply discards a significant portion of the data you mark on your ballot, whereas Approval Voting uses 100% of the data on your ballot. This explains why the 2009 IRV mayoral race in Burlington, VT elected the Progressive even though a sizable majority of voters preferred the Democrat to the Progressive.
The third factor is the accuracy of the information you mark on your ballot. IRV incentives more damaging tactical distortions of honest preferences. E.g. in that Burlington race, some Republican voters would have gotten a better result if they insincerely ranked the Democrat (their 2nd choice) in 1st place—because then the Democrat would have won instead of the Progressive (their 3rd choice).
You can see this all shake out in Bayesian Regret figures, which show the quality of a voting method as a function of all combined factors.
It’s amazing how wrong intuition usually is in the field of election theory.
A simple example is the past two Maine gubernatorial elections, where a brutally incompetent right wing climate denier and all around obstructionist won due to vote splitting between the Democrat and a left-leaning independent. But the Center for Election Science did an exit poll which showed the finish order would have reversed with Approval Voting. (The independent would have been first and the Republican last.)
IRV often fails in scenarios like this due to the “center squeeze” problem. E.g. say you have these preferences:
IRV eliminates Center first, and then pits Left vs. Right.
But 65% of the voters prefer Center to Left. and 67% prefer Center to Right. Center is indisputably the most overall preferred candidate.
Here’s a presentation I gave to the Colorado League of Women Voters last year that highlights a multitude of other advantages of Score Voting and Approval Voting over IRV.
Not really. Score Voting and Approval Voting are radically simpler to implement and explain. Borda is also much simpler. And there are even several Condorcet-compliant methods that are simpler than IRV. IRV is really the worst alternative voting system no matter how you look at it.
Clay Shentrup
Co-founder, The Center for Election Science
For the same reason people show up and vote, even though there’s an incredibly tiny chance they’ll change the election outcome. Because people like to express themselves.
Or, for the same reason lots of people for Greens and Libertarians under the present system.