The problem posed by gerrymandering is a social problem, not a math problem.
There are all kinds of geographical and legal boundaries in a country, with varying levels of affinity to them. Defining gerrymandering precisely isn't possible, it isn't even a know it when you see it thing, and in practice it's a Russell conjugation: they gerrymander, and we redistrict so that people get proper representation.
However, social problems can be solved with maths, law could require something like this to be performed after every census:
I think this has considerable advantages, the biggest being that it's objective, the next being that just looking at a Voronoi diagram makes it clear what's going on.
This essentially removes human factors from redistricting, and says you get the Congressional district you get when the population heatmap is relaxed in the simplest way it can be into the appropriate number of partitions.
Remember there's nothing about this which is 'fair', except for the part where people could perhaps be persuaded to abide by the results.
The root of the issue is that the US doesn't separate elected lawmakers from district distribution. The fact[0] elected officials can redraw their own electoral map is just beyond belief, in a tough competition with Citizen's United for the worst political bug (well... feature) in the US. Cherry on top is calling/dubbing one such major and catastrophic instance of gerrymandering "REDMAP". So brazenly rubbing it in while openly chipping at US democracy.
As long as that holds true, theoretical solutions don't really matter when those using them will pry any model into their favor. Not that a mathematical solution even makes sense for a district map that accounts of cultural ties for local representation (short of going for proportional elections, which I think is desirable with some form of local representation).
In Canada, and I presume most countries of the G7, it's an independent entity[1], which one would think goes without saying. That's not incorruptible of course, but I wager it's held it's own for over 60 years here, modern district maps are quite reasonable.
I agree that algorithmic redistricting based entirely on population density would be the best way to do strict single-member geographic representation. But even neutrally drawn districts give more power to the party that is less concentrated geographically.
Proportional representation solves this, but if you still value the notion that local communities should be able to have their collective power represented, then you need a solution that combines the two.
As mentioned elsewhere in the comments, multi-member districts with proportional representation are a good way to do this without needing to change the Constitution. The gist is, split states into a smaller number of larger districts, each of which gets multiple representatives. The representatives are allocated proportionally to the vote in each large district. So if the district has 5 representatives, party X gets 63% of the vote, and party Y gets 37% of the vote, then party X gets 3 members and party Y gets 2 members.
Political scientist Lee Drutman is actively promoting this approach, and recently founded Fix Our House (https://www.fixourhouse.org).
If I could rewrite the Constitution, I'd get rid of the rounding error by weighting the votes of individual members of Congress. E.g., if there were only two candidates in a district, candidate X got 60% of the vote, and candidate Y got 40% of the vote, send them both to Congress, with X getting 0.6 of a vote and Y getting 0.4 of a vote.
Of course all of this is better when combined with instant runoffs (aka ranked choice).
I'm glad you said this. One of the more infamous gerrymanders in the US is in Chicago --- the notorious "earmuffs". You can just look at it and see that something's hinky.
But what's really happening there is that there are in fact very well-defined Latino neighborhoods in Chicago, and the earmuffs capture a bunch of them neatly: Pilsen, Little Village, Cicero, Belmont-Cragin. If you know Chicago, you know these places, and you also know what the boundary between, say, Belmont-Cragin and North Austin is like; however artificial it looks on a map, it is a real border. That these communities are where they are is also not purely happenstance: a lot of very unfortunate social engineering took place in the early-mid-20th century to put those neighborhoods (and all the other neighborhoods) where they are now.
That's not to defend gerrymandering writ large, or even necessarily the earmuffs, but just to second your point about the complexity of the problem.
It's a funny thing, because the point of representation is that you want to group the people together by something so you can accurately represent them, but the only thing pretty much every nation has settled on is to group them by geography.
Why is that? How much of a connection does a rich white socialite in New York have with a poor custodian from an immigrant family? Sure, their representative may be trying to bring money into their district, but beyond that is it really possibly to represent both of their interests?
What if representatives could represent based on other factors, like sex or wealth or race? What would that do to the idea of representational government?
Or what if, in the online age, you could simply decide which representative represented you, no matter where they were from? Each representative could have no more than [US pop./num. districts] constituents, but they'd hail from anywhere, in some complex series of elections?
The real problem isn't even gerrymandering, it's having a district system at all. Gerrymandering is merely one of the symptoms of a district system.
Having a district system will tend to reduce the number of viable parties to two. The article doesn't even question the fact that there's only a blue and a violet party in the example. In reality, lots of people feel badly represented by even the best of those two parties. For as long as I've talked to Americans about politics, they've complained about voting for the lesser evil. Why not vote for a party you actually agree with? Because it's not viable in a two-party system.
If you've got 3 parties: Green, Yellow and Pink. 40% of the people vote Green, 30% vote yellow and 30% vote Pink. However, Green is supported all over the country, whereas Yellow and Pink are more regional, so in every district, Green gets 40% of the vote, whereas one of Yellow and Pink gets the other 60%, while the other gets none. Despite being the largest party, Green gets no votes.
Or let's say Green and Yellow each gets 40% of the vote and Pink gets 20%. All votes are distributed reasonably evenly, so about half of the districts are won by Green and half are won by Yellow. This sounds reasonable, until you realise that the Pink voters get no representation at all. It's possible to have a system where even a majority of voters wouldn't be represented, as long as the two largest parties manage to keep the smaller parties divided.
And in the US, that situation has devolved to the point that many voters don't even bother to vote for the smaller parties anymore, because they'd be throwing away their vote. And instead, they vote for candidates they disagree with because the only other viable candidate would be even worse.
Just use proportional representation. That's the only way to ensure that all political opinions get represented. It's not perfect, because nothing ever is, but at least the system is flexible, and voters can vote for completely new parties if they disagree with all existing parties, and those new parties can actually win seats.
The rural populations seems would not like a redistricting where every district is dominated by its largest population center. Essentially this voronoi type redistricting would push more power towards cities at the cost of rural areas.
Author here. Some people, who wrote comments in response to this essay, were focused on the intentional dishonesty of gerrymandering, and they wrote things like this:
"You can easily get rid of gerrymandering simply by proscribing an algorithmic process for districting."
There were several people who said this, here and on Reddit and on Hacker News. I believe they are defining gerrymandering as a deliberate and intentional act. By contrast, I was using the word "gerrymandering" to refer to the fact that any geographic boundary will establish a gap between the will of the voters and the final result of voting. So even if an algorithmic process is followed, and the damage done is accidental and unintentional, I am still referring to that as "gerrymandering." Indeed, the whole point of my essay is that if one group does its best to be fair when drawing lines on the map, and another group is evil and immoral and corrupt, and does its worst when drawing lines on a map, the damage done by both groups will be about the same, over the long-term.
I don't think objective is the word right word to describe this. It's algorithmic and deterministic. But it's not objective in the sense that it's objectively correct to group people like this. You're baking in a lot of implicit assumptions about how representation should/should not work by choosing this particular grouping algorithm.
It also doesn't remove the human factors from redistricting; humans would have to choose and approve this method. It's just laundering the gerrymandering through an algorithm to give it credibility.
But if you're the party that stands to benefit from changing redistricting laws, then the system loses credibility because you picked this method.
And if you're the party that stands to lose from changing these laws, then you're unlikely to self-sabotage yourself by doing so.
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.
> 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.)
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.
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?
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.
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.
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'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".
> You cannot have both, unless you're willing to have a legislature that has something like 5,000 people in it, which is not practical. (For instance, the USA has 50 states, so if we wanted 100 representatives from each state, the legislature would have 5,000 people in it — clearly not realistic.)
Madison said the same thing, but we didn't have contemporary communication tools. My company's Slack has over 5,000 people and works just fine. Why is it unrealistic to repeal the Permanent Apportionment Act of 1929 and increase membership in the House by an order of magnitude?
Because it might be possible to stand out in a crowd of 500 but not 5,000. And it certainly is not possible to do deals other than "this is good for everyone".
It would almost certainly end political parties. Lead to good ideas being debated openly by people who have no chance of hanging together long enough to force a direction.
Interesting, I didn't know that. I looked it up, and I really liked this Madison quote:
> In all very numerous assemblies, of whatever characters composed, passion never fails to wrest the scepter from reason. Had every Athenian citizen been a Socrates, every Athenian assembly would still have been a mob.
Increasing membership of the House has the effect of reducing the advantage less-populous states have (both in the house and in the electoral college). That might make it politically infeasible to pass.
Exactly. Why do people need representatives in the age when everyone has a voice? Education is still crucial, and not everyone can have an opinion on all matters, but the current first-past-the-post and representative system is blatantly corrupt. It's unlikely those in power would give it up easily, but we need a new system that's truly of the people, by the people, for the people.
Flux[1] was a step in the right direction, but unfortunately seems to be recently disbanded. The idea of everyone voting for issues they care about, where leaders are subject matter experts, is a sound one. It's ludicrous to think that a single person can politically represent thousands or millions of people on all matters. Modern politics is a reality show where charismatic and unscrupulous power-hungry egomaniacs compete for the top celebrity spot.
How many people in your company with 5000 slack users are the real decision makers ? I highly doubt it’s more than a few. If you had 5000 people in the legislature they would most likely still vote party line based on the decision of a few.
> the legislature would have 5,000 people in it — clearly not realistic.
Perhaps one solution would be for those 5,000 people to be entrusted with the job of picking (for each bill being considered) 100 "second-level representatives" to go forward and do the hard work of debating and evaluating proposals and reaching consensus.
These SLRs could be a mixture of subject matter experts and people good at political logrolling, so that the legislation which passed would reflect the priorities of the 5,000 representatives (and their voters) while still allowing a meaningful discussion to take place and shape the result.
"Gerrymandering is universal in any political system that has geographic units that demand representation." A unstated assumption here is a winner-takes-all election system, which is common in the US and UK, but not other countries, where parliament is proportional, and coalitions are common. It works.
A more interesting question would be how to build incentives in a winner-takes-all election system towards reducing gerrymandering, and to discuss what it would take to change away from a winner-takes-all geographical election system.
"A unstated assumption here is a winner-takes-all election system"
In terms of the math, increasing the number of representatives from each district is the same as making the districts smaller. As the essay said:
"As the districts get smaller, then the size of the shift needed (to punish the previous majority party) describes the asymptote of the tangent, converging towards the reality of what the public actually wants."
Increasing the number of representatives from each district is the same thing, as the number increases, the shift needed (to punish the previous majority party) describes the asymptote of the tangent, converging towards the reality of what the public actually wants.
But there remains some distortion, however small. You eventually get into infinitesimals that have no practical impact, but it is still there. If you have districts that only have 1,000 people each, and each district elects 1,000 representatives (everyone in the district) then the distortion disappears, but if you have districts with 1,000 people, and they only elect 999 representatives, the distortion reappears, however small that might be.
Also, from the essay:
---------------------
Some nations have fine-grained representation of land but course-grained representation of ideas
Some nations have fine-grained representation of ideas but course-grained representation of land
You cannot have both, unless you're willing to have a legislature that has something like 5,000 people in it, which is not practical. (For instance, the USA has 50 states, so if we wanted 100 representatives from each state, the legislature would have 5,000 people in it — clearly not realistic.)
Despite the author namedropping some other places, the whole piece is struggling to gain the validity it seeks due to pretending the US is the only country on earth and no one else could have figured out some solutions.
Also, the wrong use of "gerrymandering" makes me wonder how much the author actually knows about the topic.
While Germany is mentioned in the article, they don't mention the best features of the German system: two choices -- one for the candidate who will represent your district, one for the party you prefer. So you get someone who truly represents the district, and you also guarantee that the party that wins the most votes gets the most seats.
Why not solve the problem of gerrymandering by getting rid of the "winner takes all" property?
E.g. instead of having 100 districts selecting 1 representative using the winner-takes-all principle, which are very sensitive with respect to gerrymandering, make 10 districts where each one sends 10 representatives, where the distribution of representatives approximates the distribution of the submitted votes.
The crux of this is that the majority/plurality system is a winner-takes-all outcome. I would like to propose a completely fair (and totally unacceptable) Voting system:
Blues and Violets each get a single vote. On election day, everyone puts their vote into a big hat, and one of the ballots is pulled out. Whoever is on that ballot, gets a unilateral victory. This system has numerous good properties:
1. Power never falls into a state where the majority party can stay in control. Gerrymandering is a non-issue.
2. Every vote counts. Instead of a single, tie breaking vote changing the outcome of the election, each additional vote increases the odds. If there are 40% violets and 60% blues, the likelihood of a violet leader is still 40%.
3. Extreme minority votes still matter. If 99% of the population is blue, and 1% of the population is violet, violet supporters still feel their voice can be heard.
4. (edit) Political spending is massively reduced, since there is a diminishing return to spending ever more money trying to tip the 49% into being 51%.
Simple solution to get rid of gerrymandering: Create a pool of extra mandates, not assigned to a specific district. Distribute them to the parties however needed to make the overall vote proportional.
This way districts still have a representative, but there is no insentive for gerrymandering.
I'm not really an expert on the issue, but there's this suggestion that I kinda like: let people living in a district choose to vote in the adjacent districts if they want to do so. This "blurs" the arbitrary lines a bit.
Also, this is probably an unpopular opinion, but I actually kinda like gerrymandering. In the sense that I like the fact that it exposes the absurdity of counting votes and claiming that it somehow represents the voice of the people.
The outcome is clearly very dependent on what voting method is used, how the districts maps are drawn, who is eligible to vote, etc. It's all just vanity of vanities.
Why do they need to only have single-member districts? How about having fewer districts with more members per district? It seems like Fix Gerrymandering is a solution aimed at the wrong problem when the entire problem could just be eliminated.
One of the biggest problems we run into is not that there are no options, but there are so many possibilities that people who study the problem in detail end up arguing with each other rather than forming a coalition for a single reform! I also designed a system based on a simplified CPO-STV, but I keep putting off writing it up.
> Some countries, such as Israel, have purely party based parliamentary systems with no geographic representation ... Some nations have fine-grained representation of ideas but course-grained representation of land
Before putting this on a pedestal, it's worth noting that Israel:
1. Has smaller some power blocs which have non-proportional power at the parliament due to being the "swing vote".
2. Had other smaller political parties completely eliminated (voter desertion) due to ineffectiveness.
3. Had 4 elections in 2 years due to inability to form a ruling government with more than 50% of the parliament approval.
4. Had a government average run-time of 2.5 years since 1999 due to "no confidence" votes that force re-election.
Less than ideal. The American system, with its flaws, significantly more stable.
I think this article is too absolutist in its conclusions (and its interpretation of Arrow's theorem). There are definitely ways to limit the worst effects of gerrymandering. We came close with a lawsuit from Wisconsin (I forget the details) and there was hope that Justice Kennedy was waiting for that to make it to the Supreme Court, but then he decided the other way and promptly retired.
I believe that sortition is a solution to this. Being a politician should be like jury duty, there's a low but significant probability that you can lose the lottery and have to be a representative for a couple of years. The same mechanism should be uses to select peace officers also. There are certain roles that society appears to need, but that the kind of people who want to fill those roles are likely to be the worst people to do the job (i.e. anyone with a hint of megalomania). Sortition solves this problem.
This article nicely explains the problem, but is a bit defeatist about the possibility of reforms that could mitigate it. To offer one reform, allow me to present the following:
Suppose in the hypothetical country, the total number of votes cast for Blue is some number "B", and the total for Violet is "V". Further suppose that the number of seats won by Blue is "C" and the seats for Violet is "W". A non-gerrymandered map would be one where B:V is approximately C:W (or at least there is no possible change to the seats result that would make C:W any closer to B:V).
My proposed reform, therefore, is that after an election, if those ratios are not as close as possible, the electoral officers would go down the list of seats that were lost by the under-represented party (from best performance to worst) and flip the result so that the under-represented party gains representation, until the ratio can't be improved any more.
Obviously this would feel hugely unfair to the voters in the districts where the result was flipped, but they should blame the people who drew the unfair map in the first place. Indeed, the incentive would be on both parties to draw districts which were most likely to produce a balanced result that doesn't need any flipping. This method has the added bonus that it would encourage parties to appeal to voters in "safe seats" just as much as "swing seats", since everyone's vote would count towards the B:V ratio.
The maths gets a little more complicated when dealing with more than 2 parties, and with independent candidates who don't belong to any party, but I don't believe these are insurmountable obstacles. Solving the problem of gerrymandering, unrepresentative legislatures, and voter apathy with this one weird trick (that doesn't involve changing the ballot papers or counting process) seems like such a big win that the cost of a few flipped seats has to be an acceptable price to pay.
Part of why partisan gerrymandering has become such a problem is that the tech makes it easier to gerrymander.
But mostly, political party matters much more than ever before. The electorate is more ideological and the candidates for office are much more loyal to the party platform. In an era where political party mattered less than, say, geography, then partisan gerrymandering was less of a national problem.
Of course, there are several types of gerrymandering and this is just one of them. The best summary of possible solutions and their limitations I've seen was done by 538 as part of a larger gerrymandering project: https://fivethirtyeight.com/features/its-probably-not-possib...
>Congratulations, you've now created a totalitarian one-party dictatorship! The blues hold 60% of every district, so they win every district, so 100% of the national legislature is in the control of the blues, even though they only hold 60% of the vote.
I'm afraid too many people would be ok with this. There is a strong desire for "democracy" over "republic" (not a combination of the two), meaning 51% of your average voters can do whatever they want to the remaining 49%. Fundamentally, including a "republic" in the formulation means that some people get a prioritized say. I think a strong case can be made for that, but the idea itself is deeply unpopular.
I've studied redistricting and gerrymandering for 15+ years ( https://bdistricting.com ), and yes the punch line is that people don't even want geographic districts at least half the time. Sometimes you want 'your local rep' to talk to about local and regional issues, but all the argument is about ideological representation (party, culture, pet issue, etc.).
My best guess solution is a two part legislature, one house of small local districts, one house elected in at-large Proportional Representation.
Don’t some European counties like Germany have the number of delegates forced to roughly equal the total vote totals for a party? So if you have “60% violet party and 40% teal party” but the seats are split 50/50 10 closest seats move over to violet?
Theres no perfect way to solve it but the ultimate problem is social: one party is willing to gerrymander to win and the other doesn’t because they don’t have to but would if they actually saw any gain in it.
As someone living in a country that uses proportional representation for its electoral system, my first introduction to gerrymandering was this episode on the Data Skeptic podcast which I highly recommend:
https://dataskeptic.com/blog/episodes/2021/gerrymandering
[+] [-] samatman|3 years ago|reply
There are all kinds of geographical and legal boundaries in a country, with varying levels of affinity to them. Defining gerrymandering precisely isn't possible, it isn't even a know it when you see it thing, and in practice it's a Russell conjugation: they gerrymander, and we redistrict so that people get proper representation.
However, social problems can be solved with maths, law could require something like this to be performed after every census:
https://sites.math.washington.edu/~morrow/mcm/uw_1034.pdf
I think this has considerable advantages, the biggest being that it's objective, the next being that just looking at a Voronoi diagram makes it clear what's going on.
This essentially removes human factors from redistricting, and says you get the Congressional district you get when the population heatmap is relaxed in the simplest way it can be into the appropriate number of partitions.
Remember there's nothing about this which is 'fair', except for the part where people could perhaps be persuaded to abide by the results.
[+] [-] BbzzbB|3 years ago|reply
As long as that holds true, theoretical solutions don't really matter when those using them will pry any model into their favor. Not that a mathematical solution even makes sense for a district map that accounts of cultural ties for local representation (short of going for proportional elections, which I think is desirable with some form of local representation).
In Canada, and I presume most countries of the G7, it's an independent entity[1], which one would think goes without saying. That's not incorruptible of course, but I wager it's held it's own for over 60 years here, modern district maps are quite reasonable.
0: https://en.wikipedia.org/wiki/REDMAP
1: https://www.vox.com/2014/4/15/5604284/us-elections-are-rigge...
[+] [-] edbaskerville|3 years ago|reply
Proportional representation solves this, but if you still value the notion that local communities should be able to have their collective power represented, then you need a solution that combines the two.
As mentioned elsewhere in the comments, multi-member districts with proportional representation are a good way to do this without needing to change the Constitution. The gist is, split states into a smaller number of larger districts, each of which gets multiple representatives. The representatives are allocated proportionally to the vote in each large district. So if the district has 5 representatives, party X gets 63% of the vote, and party Y gets 37% of the vote, then party X gets 3 members and party Y gets 2 members.
Political scientist Lee Drutman is actively promoting this approach, and recently founded Fix Our House (https://www.fixourhouse.org).
If I could rewrite the Constitution, I'd get rid of the rounding error by weighting the votes of individual members of Congress. E.g., if there were only two candidates in a district, candidate X got 60% of the vote, and candidate Y got 40% of the vote, send them both to Congress, with X getting 0.6 of a vote and Y getting 0.4 of a vote.
Of course all of this is better when combined with instant runoffs (aka ranked choice).
[+] [-] tptacek|3 years ago|reply
I'm glad you said this. One of the more infamous gerrymanders in the US is in Chicago --- the notorious "earmuffs". You can just look at it and see that something's hinky.
But what's really happening there is that there are in fact very well-defined Latino neighborhoods in Chicago, and the earmuffs capture a bunch of them neatly: Pilsen, Little Village, Cicero, Belmont-Cragin. If you know Chicago, you know these places, and you also know what the boundary between, say, Belmont-Cragin and North Austin is like; however artificial it looks on a map, it is a real border. That these communities are where they are is also not purely happenstance: a lot of very unfortunate social engineering took place in the early-mid-20th century to put those neighborhoods (and all the other neighborhoods) where they are now.
That's not to defend gerrymandering writ large, or even necessarily the earmuffs, but just to second your point about the complexity of the problem.
[+] [-] SamBam|3 years ago|reply
Why is that? How much of a connection does a rich white socialite in New York have with a poor custodian from an immigrant family? Sure, their representative may be trying to bring money into their district, but beyond that is it really possibly to represent both of their interests?
What if representatives could represent based on other factors, like sex or wealth or race? What would that do to the idea of representational government?
Or what if, in the online age, you could simply decide which representative represented you, no matter where they were from? Each representative could have no more than [US pop./num. districts] constituents, but they'd hail from anywhere, in some complex series of elections?
Thought experiments, of course.
[+] [-] mcv|3 years ago|reply
Having a district system will tend to reduce the number of viable parties to two. The article doesn't even question the fact that there's only a blue and a violet party in the example. In reality, lots of people feel badly represented by even the best of those two parties. For as long as I've talked to Americans about politics, they've complained about voting for the lesser evil. Why not vote for a party you actually agree with? Because it's not viable in a two-party system.
If you've got 3 parties: Green, Yellow and Pink. 40% of the people vote Green, 30% vote yellow and 30% vote Pink. However, Green is supported all over the country, whereas Yellow and Pink are more regional, so in every district, Green gets 40% of the vote, whereas one of Yellow and Pink gets the other 60%, while the other gets none. Despite being the largest party, Green gets no votes.
Or let's say Green and Yellow each gets 40% of the vote and Pink gets 20%. All votes are distributed reasonably evenly, so about half of the districts are won by Green and half are won by Yellow. This sounds reasonable, until you realise that the Pink voters get no representation at all. It's possible to have a system where even a majority of voters wouldn't be represented, as long as the two largest parties manage to keep the smaller parties divided.
And in the US, that situation has devolved to the point that many voters don't even bother to vote for the smaller parties anymore, because they'd be throwing away their vote. And instead, they vote for candidates they disagree with because the only other viable candidate would be even worse.
Just use proportional representation. That's the only way to ensure that all political opinions get represented. It's not perfect, because nothing ever is, but at least the system is flexible, and voters can vote for completely new parties if they disagree with all existing parties, and those new parties can actually win seats.
[+] [-] mpalczewski|3 years ago|reply
[+] [-] lkrubner|3 years ago|reply
"You can easily get rid of gerrymandering simply by proscribing an algorithmic process for districting."
There were several people who said this, here and on Reddit and on Hacker News. I believe they are defining gerrymandering as a deliberate and intentional act. By contrast, I was using the word "gerrymandering" to refer to the fact that any geographic boundary will establish a gap between the will of the voters and the final result of voting. So even if an algorithmic process is followed, and the damage done is accidental and unintentional, I am still referring to that as "gerrymandering." Indeed, the whole point of my essay is that if one group does its best to be fair when drawing lines on the map, and another group is evil and immoral and corrupt, and does its worst when drawing lines on a map, the damage done by both groups will be about the same, over the long-term.
[+] [-] karpierz|3 years ago|reply
It also doesn't remove the human factors from redistricting; humans would have to choose and approve this method. It's just laundering the gerrymandering through an algorithm to give it credibility.
But if you're the party that stands to benefit from changing redistricting laws, then the system loses credibility because you picked this method.
And if you're the party that stands to lose from changing these laws, then you're unlikely to self-sabotage yourself by doing so.
[+] [-] onphonenow|3 years ago|reply
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.
[+] [-] dragonwriter|3 years ago|reply
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.)
[+] [-] theptip|3 years ago|reply
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.
[+] [-] twic|3 years ago|reply
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.
[+] [-] chrisseaton|3 years ago|reply
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?
[+] [-] BeFlatXIII|3 years ago|reply
[+] [-] panick21_|3 years ago|reply
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.
[+] [-] lettergram|3 years ago|reply
[+] [-] yucky|3 years ago|reply
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.
[+] [-] sidewndr46|3 years ago|reply
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.
[+] [-] drewcoo|3 years ago|reply
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.
[+] [-] abirch|3 years ago|reply
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.
[+] [-] dllthomas|3 years ago|reply
If we believe that the appropriate compromise is no one's first choice, the first thing IRV does is throw away any "appropriate compromise".
[+] [-] maxwell|3 years ago|reply
Madison said the same thing, but we didn't have contemporary communication tools. My company's Slack has over 5,000 people and works just fine. Why is it unrealistic to repeal the Permanent Apportionment Act of 1929 and increase membership in the House by an order of magnitude?
[+] [-] lifeisstillgood|3 years ago|reply
It would almost certainly end political parties. Lead to good ideas being debated openly by people who have no chance of hanging together long enough to force a direction.
My god man what are you thinking !
(/s)
[+] [-] cato_the_elder|3 years ago|reply
Interesting, I didn't know that. I looked it up, and I really liked this Madison quote:
> In all very numerous assemblies, of whatever characters composed, passion never fails to wrest the scepter from reason. Had every Athenian citizen been a Socrates, every Athenian assembly would still have been a mob.
(from https://en.m.wikipedia.org/wiki/Federalist_No._55)
[+] [-] aidenn0|3 years ago|reply
[+] [-] imiric|3 years ago|reply
Flux[1] was a step in the right direction, but unfortunately seems to be recently disbanded. The idea of everyone voting for issues they care about, where leaders are subject matter experts, is a sound one. It's ludicrous to think that a single person can politically represent thousands or millions of people on all matters. Modern politics is a reality show where charismatic and unscrupulous power-hungry egomaniacs compete for the top celebrity spot.
[1]: https://voteflux.org/
[+] [-] spaetzleesser|3 years ago|reply
[+] [-] dane-pgp|3 years ago|reply
Perhaps one solution would be for those 5,000 people to be entrusted with the job of picking (for each bill being considered) 100 "second-level representatives" to go forward and do the hard work of debating and evaluating proposals and reaching consensus.
These SLRs could be a mixture of subject matter experts and people good at political logrolling, so that the legislation which passed would reflect the priorities of the 5,000 representatives (and their voters) while still allowing a meaningful discussion to take place and shape the result.
[+] [-] lkrubner|3 years ago|reply
[+] [-] photochemsyn|3 years ago|reply
[+] [-] david_draco|3 years ago|reply
A more interesting question would be how to build incentives in a winner-takes-all election system towards reducing gerrymandering, and to discuss what it would take to change away from a winner-takes-all geographical election system.
[+] [-] lkrubner|3 years ago|reply
In terms of the math, increasing the number of representatives from each district is the same as making the districts smaller. As the essay said:
"As the districts get smaller, then the size of the shift needed (to punish the previous majority party) describes the asymptote of the tangent, converging towards the reality of what the public actually wants."
Increasing the number of representatives from each district is the same thing, as the number increases, the shift needed (to punish the previous majority party) describes the asymptote of the tangent, converging towards the reality of what the public actually wants.
But there remains some distortion, however small. You eventually get into infinitesimals that have no practical impact, but it is still there. If you have districts that only have 1,000 people each, and each district elects 1,000 representatives (everyone in the district) then the distortion disappears, but if you have districts with 1,000 people, and they only elect 999 representatives, the distortion reappears, however small that might be.
Also, from the essay:
---------------------
Some nations have fine-grained representation of land but course-grained representation of ideas
Some nations have fine-grained representation of ideas but course-grained representation of land
You cannot have both, unless you're willing to have a legislature that has something like 5,000 people in it, which is not practical. (For instance, the USA has 50 states, so if we wanted 100 representatives from each state, the legislature would have 5,000 people in it — clearly not realistic.)
[+] [-] oxnrtr|3 years ago|reply
Despite the author namedropping some other places, the whole piece is struggling to gain the validity it seeks due to pretending the US is the only country on earth and no one else could have figured out some solutions.
Also, the wrong use of "gerrymandering" makes me wonder how much the author actually knows about the topic.
[+] [-] not2b|3 years ago|reply
[+] [-] q-big|3 years ago|reply
E.g. instead of having 100 districts selecting 1 representative using the winner-takes-all principle, which are very sensitive with respect to gerrymandering, make 10 districts where each one sends 10 representatives, where the distribution of representatives approximates the distribution of the submitted votes.
This should decrease the problem a lot.
[+] [-] bumper_crop|3 years ago|reply
Blues and Violets each get a single vote. On election day, everyone puts their vote into a big hat, and one of the ballots is pulled out. Whoever is on that ballot, gets a unilateral victory. This system has numerous good properties:
1. Power never falls into a state where the majority party can stay in control. Gerrymandering is a non-issue.
2. Every vote counts. Instead of a single, tie breaking vote changing the outcome of the election, each additional vote increases the odds. If there are 40% violets and 60% blues, the likelihood of a violet leader is still 40%.
3. Extreme minority votes still matter. If 99% of the population is blue, and 1% of the population is violet, violet supporters still feel their voice can be heard.
4. (edit) Political spending is massively reduced, since there is a diminishing return to spending ever more money trying to tip the 49% into being 51%.
[+] [-] thomasahle|3 years ago|reply
This way districts still have a representative, but there is no insentive for gerrymandering.
This is a commonly known as Overhang seats in MMP: https://en.m.wikipedia.org/wiki/Overhang_seat
[+] [-] cato_the_elder|3 years ago|reply
Also, this is probably an unpopular opinion, but I actually kinda like gerrymandering. In the sense that I like the fact that it exposes the absurdity of counting votes and claiming that it somehow represents the voice of the people.
The outcome is clearly very dependent on what voting method is used, how the districts maps are drawn, who is eligible to vote, etc. It's all just vanity of vanities.
[+] [-] no_protocol|3 years ago|reply
[+] [-] scythe|3 years ago|reply
Australian five-member STV districts beg to differ. As does MMP:
http://en.wikipedia.org/wiki/Mixed-member_proportional_repre...
And DMP:
https://en.m.wikipedia.org/wiki/Dual-member_proportional_rep...
And PR-OAC:
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2778867
One of the biggest problems we run into is not that there are no options, but there are so many possibilities that people who study the problem in detail end up arguing with each other rather than forming a coalition for a single reform! I also designed a system based on a simplified CPO-STV, but I keep putting off writing it up.
[+] [-] bironran|3 years ago|reply
Before putting this on a pedestal, it's worth noting that Israel:
Less than ideal. The American system, with its flaws, significantly more stable.[+] [-] tunesmith|3 years ago|reply
EDIT: I'm thinking of the Efficiency Gap: https://en.wikipedia.org/wiki/Wasted_vote#Efficiency_gap
[+] [-] pacaro|3 years ago|reply
[+] [-] dane-pgp|3 years ago|reply
Suppose in the hypothetical country, the total number of votes cast for Blue is some number "B", and the total for Violet is "V". Further suppose that the number of seats won by Blue is "C" and the seats for Violet is "W". A non-gerrymandered map would be one where B:V is approximately C:W (or at least there is no possible change to the seats result that would make C:W any closer to B:V).
My proposed reform, therefore, is that after an election, if those ratios are not as close as possible, the electoral officers would go down the list of seats that were lost by the under-represented party (from best performance to worst) and flip the result so that the under-represented party gains representation, until the ratio can't be improved any more.
Obviously this would feel hugely unfair to the voters in the districts where the result was flipped, but they should blame the people who drew the unfair map in the first place. Indeed, the incentive would be on both parties to draw districts which were most likely to produce a balanced result that doesn't need any flipping. This method has the added bonus that it would encourage parties to appeal to voters in "safe seats" just as much as "swing seats", since everyone's vote would count towards the B:V ratio.
The maths gets a little more complicated when dealing with more than 2 parties, and with independent candidates who don't belong to any party, but I don't believe these are insurmountable obstacles. Solving the problem of gerrymandering, unrepresentative legislatures, and voter apathy with this one weird trick (that doesn't involve changing the ballot papers or counting process) seems like such a big win that the cost of a few flipped seats has to be an acceptable price to pay.
[+] [-] elicash|3 years ago|reply
But mostly, political party matters much more than ever before. The electorate is more ideological and the candidates for office are much more loyal to the party platform. In an era where political party mattered less than, say, geography, then partisan gerrymandering was less of a national problem.
Of course, there are several types of gerrymandering and this is just one of them. The best summary of possible solutions and their limitations I've seen was done by 538 as part of a larger gerrymandering project: https://fivethirtyeight.com/features/its-probably-not-possib...
[+] [-] daenz|3 years ago|reply
I'm afraid too many people would be ok with this. There is a strong desire for "democracy" over "republic" (not a combination of the two), meaning 51% of your average voters can do whatever they want to the remaining 49%. Fundamentally, including a "republic" in the formulation means that some people get a prioritized say. I think a strong case can be made for that, but the idea itself is deeply unpopular.
[+] [-] brianolson|3 years ago|reply
My best guess solution is a two part legislature, one house of small local districts, one house elected in at-large Proportional Representation.
[+] [-] marricks|3 years ago|reply
Theres no perfect way to solve it but the ultimate problem is social: one party is willing to gerrymander to win and the other doesn’t because they don’t have to but would if they actually saw any gain in it.
[+] [-] almog|3 years ago|reply