Hacker News Reader

Comment by onphonenow | Hacker News Reader

Very quick solutions that at least drive politics towards the middle vs extremes:

Open primaries with ranked choice voting.

Top two to general.

Done.

The current model, in a "safe" dem seat (45% R / 55% D) whoever wins D primary wins everything. This means the D primary (with 20% of voters - most activist) are deciding things. Same on the right.

With open primaries, you will get 2 D's potentially in general, but now the R vote actually matters. This drives towards the middle.

Additionally RCV in the primary (ie pick 3 in order) also allows votes for less well known folks without as much wasted voting.

Another method is RCV with a multi-seat election. So you rate 5 candidates and their are 5 winners in the larger geo area. This tends to give even groups with let's say 20% makeup of voter pool a seat at the table. Traditional winner takes all does not.

View on HN

> Very quick solutions that at least drive politics towards the middle vs extremes

This is a very good goal if you want to replace duopoly which alienates a large number of people with monopoly which alienates a larger number of people.

What we need is almost the opposite: a system that, to the extent possible without compromising things like manageable legislature size and personal electoral accountability, provides proportional representation so that governing compromises are forced out in the open between parties in government, rather than done behind the scenes within parties so as to attempt to attain a minimal winning coalition leveraging limited substantive choices in the general election.

(Using Single Transferrable Vote, or a similar system, in small multimember districts, say 5 members per district, would do this quite well.)

Can you say why you were prefer STV over simple approval voting where the top 5 vote getters win? Substantial distortions are still possible with STV, and it is complicated. By contrast "50 candidates are running, the voters get 50 votes, the top 5 vote getters win" will have distortions similar to STV, but the math is simple and everyone can understand it. What benefit does STV have?

Score voting is far better then approval or ranked choice.

Giving every candidate 1-5 or 1-10 points like Amazon reviews is by far the best system.

But Netflix tried score voting for years, with tens of millions of users, and in the end they gave up on score voting and switched to approval voting.

That might be the right system to make voting on lots of media practical threw terrible input devices but when we are talking about a political system that decides how to use nuclear weapons, the effort is worth it.

I think Top-2 is a dead-end. It bakes in the assumption of parties as the ultimate power brokers, and further entrenches FPTP.

Even simpler option: implement STV. Then primaries are optional as you can essentially bake the primary candidates into the main election if you like.

I think it’s quite repugnant to offer Republicans the choice of two Democrats to vote for. (I’m not a Republican). Top-2 admits that FPTP with party primaries is broken, but declines to actually fix the problem beyond a band-aid.

This would be an improvement, but i think it illustrates the paucity of thinking in American politics.

You shouldn't be using single-member districts at all. It's basically impossible to represent people fairly with them. You should either have proportional election of entire legislatures (or state delegations to federal ones), or at least proportional election of a smaller number of multi-member districts.

> Open primaries

I don't get the idea of open primaries - why should someone who is not a member of a party get to vote on who the party wants to represent them?

Imagine there was a Women's Equality Party party (there is in the UK.) Should a large number of men, with no interests in women's equality, be able to come along and tell the Women's Equality Party who will represent them in an election?

edit: was describing a jungle primary

If that is what GP meant, then they used the wrong term. The term usually used is "Blanket Primary" or "Jungle Primary" (Jungle Primary is a specific type of blanket primary) for what you describe. An open primary allows people to vote regardless of party affiliation; the states which have these may not even ask you for party affiliation when you register to vote (Virginia, where I first voted does not, for example).

It does make an aggregate difference because few people vote in primaries anyways, and even fewer will change their party affiliation every year, making raiding much harder.

That's a California-style jungle primary, not an open primary, and it's very dubious that it promotes centrism.

The California jungle primaries destroyed almost all of the right wing wingnuts in California. That's practically centrist by tautology.

However, the more interesting cases are actually where there have been two Democratic candidates, and, yes, much to my irritation, it really does seem to promote "centrism". See: Harris vs Sanchez for US Senate, for example.

Now, whether "centrism" is a good thing is up for debate as history shows that the centrist position is almost always the wrong position over the long run.

> The California jungle primaries destroyed almost all of the right wing wingnuts in California

No, they didn't.

> However, the more interesting cases are actually where there have been two Democratic candidates, and, yes, much to my irritation, it really does seem to promote "centrism".

Almost invariably, the candidate thar had the most institutional support (usually, a more centrist one) in the primary wins the general in that case, same as they likely would have in the prior system where the institutional support would have won them the party nomination. The difference is that they campaign against another (usually more leftist) Democrat in the general.

California used to have actual moderate republicans. Yes even in the Bay Area. Over time those guys retired. The voters moved left a little. And the Republicans followed the national party moving way to the right. The result was the Republicans inability to field viable candidates against Democrats.

The jungle primary was a response the loss of a viable opposition party.

Got to say after Feinstein won against de Leon in 2018 I'm kinda done with hearing conservatives bitch about her.

> anybody can register for any party

Presumably only if they'll have you? People often get ejected from political parties in the UK if they for some reason are being destructive (in someone's opinion.)

Not in Alaska, anyone can assert any affiliation in our primary: "If a candidate is registered as affiliated with a political party or group, it does not imply the candidate is nominated by, endorsed by, approved of, or associated with that particular party or group."

>Presumably only if they'll have you?

Not in the US. Primaries are just the method that political parties use to determine who they should run in the actual election. You do not ever need party affiliation to vote in the actual election, but some state primaries are run in a manner that require you to register with the party if you want to vote in their primary.

Primaries are typically held months in advance of the actual election.

Given that parties don't restrict membership in the US, in places that are majority one party, you effectively get an open primary. I'm registered as $MAJORITY_PARTY_IN_MY_DISTRICT because the primary election is essentially the only one that matters.

I am someplace that has open primaries and nothing like you are concerned about transpires. The reason is that if you participate in one primary you lose the ability to participate in the other. In some sense the threat of each side doing it keeps people from doing it.

Also you can't really in keep people from registering for a given party, so enforcing a non open primary isn't feasible in practice.

Alaska switched to a non-partisan top 4 RCV primary. This seems much more likely type of primary to actually benefit moderate candidates and voters. We've had moderate candidates get primaried and have to run and win write-in campaigns to stay in office. This means that reasonably popular moderate candidates will at least always get to run in the general election.

RCV has been passing in referendums here, which is changing some of the primaries, but at a very local level.

The point is that the party has no intrinsic right to a slot in the general election.

The WEP can put whoever they wish up to represent them, selected however they wish, in a process funded by their party resources.

Then the primary is a voting population wide decision to winnow the field to a reasonable final competition.

A state funded primary that is locked to a party selection process is an interesting use of public funds.

I can think of two reasons -

1 - Most voters are not a Republican or Democrat. So they're stuck choosing between what those at political extremes want at election time.

2 - Some cities/states will always vote R or D, pretty reliably. Choosing someone you dislike less has its advantages in such areas.

The current iteration of primaries, and the government paying for them, got going in the USA during the 1960s. It was the start of the populist revolt, and part of the goal was to weaken the political parties. Progressive reformers were motivated by the outcomes of struggles such as Fanny Lou Hamer.

https://en.wikipedia.org/wiki/Fannie_Lou_Hamer

Civil Rights leaders wondered why ultra-racist white men got to represent the Democratic Party in Mississippi, when someone like Fanny Lou Hamer seemed to have broad based support in her district. Why were ultra-racist Democrats able to make a backroom deal and nominate some staunch segregationist, instead of allowing someone to stand for the Democrats, chosen democratically?

The goal of the primaries was to weaken the political parties, which is why you might find them confusing. At the time, a number of scholars pointed out that weakening the parties would increase the influence of money, and also allow untested demagogues to step forward and become the leaders of the weakened parties. These concerns seem to have been borne out by events.

In a voting system that is not insane a party doesn't get punished for running more then 1 person.

So you could just have 3 people from the WEP run and then voters decide.

If the party has enough power over its member to force only 1 candidate, then they can do that. However once parties get larger enforcing that is difficult and other people will just run.

Why not simply run something other than FPTP in the general and eliminate primaries altogether?

Ranked choice voting in general is a real bad idea. Its better then first past the post but that's it.

If you are gone change the voting system, why not just go for the much better score voting, or Star vote.

RCV leads to incredibly very strange and non-intuitive outcomes in some situations. Score voting beats it in pretty much every way while also being far, far simpler for people to actually use.

In general you need to eliminate all per-elections and make districts larger as well.

Then you are starting to approach something actually good.

Easiest thing is to ban party affiliation in ads and on ballots

Doesn’t work very well with judges, which are elected in many states.

I like this commenters suggestion of ranked voting. Would lead to more centrist candidates.

For example, i believe that’s what we saw play out in France, essentially. Center-left Macron got most votes. Far-right Le Pen got second most votes. And far-left candidate got third most votes. Once it was down to the top 2, Macron handily won.

> “Far-right Le Pen”

> “Center-left Macron”

That media / framing is exactly why the election method doesn’t work as well as banning affiliation. Le Pen is not really “far right” by most standards, but media labels her in that position and can get away with it due to name and party recognition.

You can still do both, change the election method and ban affiliation. Probably would be the best of both worlds.

>Open primaries with ranked choice voting.

Primaries aren't part of the official voting process though. They're entirely defined by parties and in fact parties can completely do away with them if they want. Primaries aren't actual elections and election laws don't really apply so there is no mechanism to force them to be open.

The solution it would seem, is to do away with political parties completely. Just like George Washington suggested from day one.

How do you define parties in such a way that you can effectively outlaw them? And would that not make fundraising even more important to individual candidates, when there’s no party mechanism to provide support?

>How do you define parties in such a way that you can effectively outlaw them?

To start, you remove party designations from the ballot and any other official voting material. Then you block local governments from being able to assist political parties with primaries/caucuses/conventions - no governmental support for any of the party infrastructure.

People want to have informal parties (ie Tea Party, The Squad, Freedom Coalition, Congressional Black Caucus etc), you can't stop that without infringing on free speech. But government isn't required to support all of the party infrastructure outside of the general election ballot access rules. And with no political party designation on ballot, anybody who meets requirements could gain access, which means no need to fall in line with a party platform. People can vote their conscience.

Enshrining the two party system in law is something that even the current power system in the US hasn't managed. By doing "top two to general" you'd be converting a de-facto duopoly into an actual duopoly.

At that point we might as well just skip the entire election process to save money and let the parties work out a power sharing agreement where it alternates every so many years.

I don't think so – France uses this exact model, and their party system has just in the last 15 years or so transitioned out both of the former top two parties (PS and LR).

If anything, I would expect this model to reduce duopoly. An unpopular republican might be overtaken by a candidate from the libertarians in the first round, giving them a shot at victory which they could hardly get if the ghost of "wasted votes" on third-parties was haunting them.

> Enshrining the two party system in law is something that even the current power system in the US hasn't managed. By doing "top two to general" you'd be converting a de-facto duopoly into an actual duopoly.

Just the opposite, the top two candidates, regardless of what party they're from will appear on the ballot, this would allow people to vote for a 3rd party candidate in the primary without worrying about "wasting" their vote. If the 3rd party candidate is one of the top two they'll appear in the general, if not that voter can still pick the lesser of two evils in the general.

The US government at all levels is controlled by two parties. While neither side is going to just hand over control to the other, they certainly aren't going to vote in a system that would allow a 3rd party to emerge.

There is no way such a system would ever be allowed to produce a ballot without a Republican and Democrat on it. If you're not on the ballot, you don't get elected. Full stop. There would be all sorts of bizarre measures and "tests" added to such a system before it was ever in place to make sure that only the existing parties would ever be on the ballot.

I guess we could just throw our hands up and say nothing will ever change, but I think things could change, and a great first step would be a grassroots push to adopt voting systems which make it easier for 3rd parties to break in.

If you choose to do that, it is your own decision. I'm not going to sit here and pretend that "top two to general" will become anything other than "Republican or Democrat" in America. I was already coerced into reading textbooks by the state that told me the US has a "two party system" and that when you vote you can choose between Republican or Democrat. This is obviously propaganda, but I'm not going to support a move to a system that makes propaganda a reality. It's incredibly dystopian.

I, for one, think that extremely D and extremely D and even somewhere in the middle are all equally problematic.

As long as 3rd parties are not electable, we're stuck with some right wing, corporate-backed politician regardless of who wins.

All the talk of gerrymandering is just a partisan distraction.

It's also Democratic Party excuse-making. They act as if Republicans invented and have a greater propensity to gerrymander than Democrats, when they both aggressively and energetically gerrymander when they're in the position to. The constant discussion of it is to distract from Democrats being completely wiped out in state and local elections over the past 20 years, giving exclusive gerrymandering power to Republicans.

You are right, Democrats are catching on, and the house republican bias will be much smaller this year than it has been in the past. That being said, Gerrymandering still cause major issues in many state elections. Wisconsin for example had well more than 50% of votes go to democrats, yet republicans still hold a super majority in their legislature, making the dem governor powerless. North carolina has similar issues. I'm curious which state you think has gerrymandered for the dems so successfully that even when 55% of the votes go to dems republicans are still able to pass any law they want.

As soon as 3rd party candidates become viable, they will become owned by special interests also.

And calling all politicians "right wing" is pretty unclear on the concept.

Perhaps they are when compared to your opinions, but if you can't see daylight between your options you have been manipulated to extreme cynicism by corporate forces that benefit from your disengagement.

I'm with you on the ranked choice voting, but I would say that the general is the duopoly. This can enable more viable parties and the death to the duopoly.

And having more representatives (districts) will help immensely as well. The current limit is capped by statute and can be changed, this would especially help counter the over-representation of unpopulated states.

RCV encompasses a lot of different systems, but the most common instantiation is something resembling instant runoff, and I am skeptical that it drives towards the middle.

If we believe that the appropriate compromise is no one's first choice, the first thing IRV does is throw away any "appropriate compromise".

Yep. Actually the behavior of IRV is extremely unintuitive (despite how nice it sounds). Even using very simple models that assume every voter has political opinions completely described by a 2d political compass, prefers candidates that are closer to them on the compass, and that voters are normally distributed around a central point, you get very strange behavior! Here are some diagrams that illustrate the effect: https://rangevoting.org/IEVS/Pictures.html

interestingly, IRV is not even monotonic! It's easy to construct a situation where converting people to have the same ranking as you causes you to get a candidate you like less. (The concave parts of the voronoi diagrams describe areas where this happens.) If you're curious about this, a few years back some friends and I tried to make a video describing this phenomenon: [0]. (Our editing and presentation skills are nothing to write home about but I think the video is interesting.)

[0]: https://www.youtube.com/watch?v=0m7vtt9jXpM

Here is another, interactive explanation of this phenomenon in IRV, as well as some alternatives that do not suffer this issue: https://ncase.me/ballot/

I don't understand why Ranked is so popular. Range voting is just easier and better in every single way.

> better in every single way.

Range can't satisfy Condorcet (that is, if there is a candidate that would win every head-to-head, that candidate should win the election).

Of course, the most common approaches to RCV also fail that criterion, but it's potentially a way some ranked methods are better than range methods.

> Range can't satisfy Condorcet

That is not necessary a desirable feature. If that is desirable or not depends on your value system of what elections should achieve. What is 'fair' or 'correct' is not universal.

Granted, but it seems a reasonable enough thing to ask for that "better in every single way" deserves qualification.

Most US districts are not competitive. E.g., there's an almost zero chance that AOC, MTG, Cawthorne, etc. loses in the general election. You have a higher probability of losing a primary than a general.

If a moderate of their party runs, it'd be easy for their own members to vote for the moderate as #1 ahead then #2 the more extreme person, and not for the awful evil other party person. The opposition can also vote for the moderate of the dominant party, #1, their candidate #2, and not for the aweful extremist.

I think that's true, and pushes things more extreme, but is pushed against (incompletely) by jungle primaries and to a lesser extent by open primaries; and ... I don't see the relationship with RCV, which was all I was discussing. Did you mean to respond one level up?

I'm not disagreeing with you about the instant runoffs for competitive elections. Then the more moderate could get squeezed out.

The relationship with RCV for non-competitive races (e.g., those districts that are gerrymandered or naturally not competitive which is the majority in the USA) is the RCV can easily address the extremes versus moderates. Moderate candidates don't run because they won't get support from their party and are perceived to be stealing votes. E.g., If you're not voting for $EXTREMIST you're voting for the $THEM. RCV lets you vote for the moderate AND your party's extremist in order to thwart the other side.

"If we believe that the appropriate compromise is no one's first choice"

The winner in RCV will always be someone's first choice.

No, some methods of counting rank choice ballots will sometimes pick a candidate who was no one's first choice. In particular, this must be so for any Condorcet method, because the Condorcet winner might be no one's first choice (imagine a setting where 1/3 of voters want candidate A then D, 1/3 want B then D, and 1/3 want C then D; 2/3 of voters would vote for D whether run head-to-head against A, B, or C, and yet D got no first place votes).

In instant runoff it is certainly true that the winner is always someone's first choice, and that was my point - sometimes the winner the method selects should not be the winner.

Reply on news.ycombinator.com