The natural answer would seem to be to use a shortest-splitline program which incorporates and prioritizes meaningful boundaries other than the overall state border. Like, any incorporated city with a population smaller than a district must be kept whole within a district. A city larger than a district but smaller than two shall be split only once, along a natural or manmade physical line that demarcates recognizable parts of the city (like a river, or a surface rail line or highway). Make a whole table of ranked if-then criteria, and then apply the bare mathematical splitlines at the end of that, to gather up as many scattered small communities as needed, to bring districts up to correct and equal sizes.
There must have been at least 350,000 people in these districts before the take-over or the districts wouldn’t have existed. So the people who took over the district must be more than 350,000. They’re as entitled to the district as the original people were. And the original 350,000 people can just move to another district - perhaps the one the people who took over their district vacated.
The point is that people can virtually never be stuck in a district where they feel their political views are being ignored. If their current district doesn’t fit them, they can just “move” to another one they’re more compatible with. You’ll never have a liberal stuck in a conservative district or a conservative stuck in a liberal district. Groups like the libertarians or the socialists or the greens can ignore their geographical diffusion and unite together. The only people who would be unrepresented are those who are so far out on the political fringe that they can’t collectively gather 350,000 together in the entire national population.
Why would the other 50% agree to this?
Or you can have it be an equal number from each party (but not active legislators or candidates), plus a chair/moderator they themselves select, who can break ties only if the rest are still hung after a number of tries.
It’s unbiased, though. Whereas gerrymandered boundaries are chosen to keep the incumbents in power, even a poor-quality algorithmic division won’t intentionally favor one candidate over another.
At any rate, it’s not a problem to use a more sophisticated algorithm. Computing power is cheap, and the results don’t have to be provably optimal, just provably reasonable.
I don’t see how you could have residency rules in a system like this.
It would be very hard to organize this. Let’s say John Smith is the elected representative of District 311. So he must have a few hundred thousand people who support him.
But there are also millions of people who despise Smith. Some of them could theoretically organize together to collectively register with District 311 and vote for somebody else on Election Day. But so what? Smith’s supporters would just move out of District 311 as his opponents moved in. They’d simply move to a different district and vote for Smith again.
As long as Smith retains a big enough group of supporters and those supporters are free to move around as they wish, then Smith’s supporters will always be able to gather themselves together and elect Smith. Smith’s opponents will never be able to create a situation where they outnumber Smith’s supporters because people can always leave a situation.
That’s an issue. Suppose Smith becomes so popular he attracts 3,000,000 supporters to his district. His popularity is now counter-productive because Smith is still only one vote in Congress. Some of his supporters would be better off moving to other districts and voting for other Congressmen who share Smith’s views. The practical advantages of splitting off like this would keep districts from becoming too overpopulated.
If there is a 10% minority population in a district it is true that the representative might ignore them especially of the majority thinks this is good. But there are a lot of things that different coalitions of voters want - it is not like the interests of the 51% are disjoint from those of the 49%. My daughter interned for a state legislator one year, and part of her job is calling to get stuff constituents wanted from state government done. I don’t think anyone asked what party the caller was from.
In any case, since I assume these virtual seats will be contested, there will still be millions of people represented by a congressman they didn’t vote for.
I just thought of something else - if I believed in abortion rights, I might want to join the right to life virtual district. And vice versa. It could be a mess.
One of the reasons we don’t have tax reform, which everyone supposedly supports, is that those who care about a specific deduction have a louder voice than those who don’t. Unless the for and against virtual districts are balanced, there will be a lot of “vote for this thing my constituents care about and yours are apathetic about and I’ll do the same for you.” Is that going to lead to better legislation? Not at all certain.
That’s another great argument against. Say one guy, named Ronald Crump, had 3 million supporters. He’s not going to abandon them, he’ll find some lackeys (Gris Cristy, for instance) who he will endorse if they swear allegiance to him. He then goes to Congress with a bloc. Today there may be influential Congresspeople, but they are pretty much limited to one area.
We might get the equivalent of warlords in Congress.
So-called unbiased redistricting is done in California by, what else, a committee. Supposedly non-partisan they have apparently made a fair job of it. Does anyone know how this works?
California, where Democrats got 59% of the vote and 73% of the seats. That’s your idea of a “fair job of it”?
Isn’t there an easy solution to this?
Everyone votes for a candidate. Every candidate is registered with a party.
First step is to count up the total number of votes for each party. Seats in the legislature are apportioned based on total votes cast for that party.
Then sort the candidates for each party by the number of people who voted for them, and hand out seats, starting with the candidate who got the most votes.
A more sophisticated version of this could allow for sub-parties… ie, there are two wings of the democratic party, one led by Bernie and one led by Hillary. Voting for either of them adds to the total count of seats the democrats get in the legislature, but in addition, voting for Bernie doesn’t just put him in the legislature, it also (if he gets sufficient votes) put various of his compatriots put into the legislature as well.
(Might as well also let people vote for a party without voting for a specific candidate, if they want to…)
I’ve always thought that letting the math decide is the most fair way. It takes politics completely out of it.
It’s a complicated topic, but my understanding is that in general, a truly “fair” system will often magnify the influence of the majority. For instance, consider two imaginary extremes:
(1) The entire state is evenly gray. That is, you can subdivide down to the block-by-block level, and every area you examine will be 59% Dem
(2) The entire state is made up of large, distinct areas which are 100% Dem or 100% Rep. In fact, there’s no reasonable way to divide things up other than to make these blocks the districts
In the first extreme, any fair system will end up with 100% democratic seats. In the second extreme, a fair system will end up with 59% democratic seats.
In the real world, we’re somewhere between those two extremes, so we’d expect to end up somewhere between 59% and 100% democratic seats.
(Which is not to say that the people who draw boundaries in California are in fact fair and non-partisan…)
Sure, and why not hold Black Sabbaths and draw pentagrams if we’re going to resort to TOOLS OF THE DEVIL!!
Nah. We already tried that.
I’d start by using zip codes. A Congresscritter might represent one or more, or only half of one based on population density. The USPS ain’t all that quick to change zip codes.
Yes. 2 USC 2(c).
Well, no. Only if you adopt an axiom that a truly fair system has to have single-member districts.
Single-member districts will always distort representation since, regardless of how votes are apportioned between two or more parties, 100% of the representation will go to one party and 0% to any of the others. And when you say that such a system “will often magnify the influence of the majority”, you are understating the phenonomenon; it will often magnify the influence of a party which fails to get a majority. In fact, this is quite common; in the most recent UK general election the Conservative Party got only 36.8% of the national vote - practically 2 out of 3 voters prefer someone else - yet got a majority of the seats. At the election before that, they got 36.4% of the votes, but 47.1% of the seats. And, at the election before that, the Labour Party got 35.2% of the votes, and 55.2% of the seats.
Of course, some consider this a feature, not a bug; the system tends to deliver a government with a strong majority in Parliament, which makes for stability and effectiveness, albeit at the price of delivering a government that most people do not want. You may take the view - I do myself - that this isn’t fair, but the OP doesn’t ask about fairness in general; he specifically focusses on fairness “by doing away with gerrymandering”. And it has to be acknowledged that this distorting feature of a system of single-member constituencies really does magnify the incentive to engage in gerrymandering. So, if you are committed to single-member constituencies, then it’s all the more necessary to take steps to prevent or discourage gerrymandering.
And I think you can do that easily enough, if the will is there. Assign the business of drawing district boundaries to an independent commission, chaired by a judge or ombudsman, and not containing political figures, and give them parameters for the drawing of constituency boundaries which refer to physical and social geography and traditional community identities, but make no reference to party support. This works reasonably well in other countries; there’s no reason why it couldn’t work in the US.
In Little Nemo’s system, why assume that there are set districts to begin with? You just decide how many people you want each representative to represent, and then any candidate who can get that number of people to affiliate with them is a representative. In this situation, there wouldn’t be anyone “stuck in the district” of a representative who doesn’t agree with them, because they would never have voted for him, and so never joined that district (which came into existence with that candidate).
I think you’re talking about something other than what I’m talking about. For this particular comment, I was speaking purely in the context of the current USA system, with states divided into a geographic districts, and two major parties, and winner-take-all within districts. In particular, I was responding to this comment from HurricaneDitka:
He was using the disparity between votes cast and seats as prima facie evidence that the ostensibly non-partisan districting in California was unfair, and I was disputing that.