21

According to Wikipedia's article on Gerrymandering:

In the process of setting electoral districts, gerrymandering is a practice that attempts to establish a political advantage for a particular party or group by manipulating geographic boundaries to create partisan or incumbent-protected districts.

Critics of Gerrymandering might argue that Gerrymandering is overly supportive of those in power, or in some cases the majority power. For instance, in Democratic controlled US states, districts might be arranged or designed to cement Democratic control of state politics. Likewise in my home town Austin, most of the districts stretch hundreds of miles to the borders of Mexico and other cities in order to disperse Democratic voters and to ensure Republican dominance of state politics.

The issue I'm concerned with is, how does one fairly design a system that offers neither party, what many view as an unfair advantage in elections?

Another problem is that populations and demographics are constantly changing and there is a legitimate need to reevaluate voting maps.

So are there any alternate viable voting district planning strategy processes that can avoid the problems of Gerrymandering?

What are considered some of the fairest strategies?

  • <comments removed> Please do not use comments to answer these questions. Comments are here to help improve the questions. Please avoid turning comment threads into your own personal chat rooms. Thanks. – Robert Cartaino Dec 19 '12 at 18:46
  • Related resource: Redistricting Game – Bobson Nov 14 '13 at 14:31
  • I recently wrote a paper on reverse gerrymandering that I think provides a realistic alternative to the current political process. I would be happy to send a copy of the paper to anyone who requests a copy. My email is pwf@northwestern.edu – user2370 Dec 23 '13 at 21:11
  • I'm a bit confused as to the question: do the solutions here need to involve geographic districts, or not? – Avi Dec 23 '13 at 22:29
16

One alternative is to let a totally impartial computer decide, based purely on census data and geography, with no details about the political (or other) makeup of the population. Brian Olsen's open source census-based B-districting algorithm aims for:

Across all districts and all people, The best district map is the one where people have the lowest average distance to the center of their district.

Comparative pictures, using North Carolina's House of Representatives districting. North Carolina using current method:

Picture of NC's jaggy rambling districts.

B-districting, based on 2010 census:

B-district NC picture, resembles patio stones.


(As a satirical gesture in lieu of posting to meta Brythan felt inspired to add the following interesting critique of B-districting. Unfortunately the critique relies on a usage of the term "community" that seems practically synonymous with current two-party districting, and includes an alphabetic diagram that I'm so far unable to decipher. -- agc)

Of course, this geometric compactness breaks geographic compactness, where neighbors who share a community are put in the same district. And of course they aren't necessarily fair. Geometric compactness does not protect minority representation. This shifts the battle from drawing the districts to picking the algorithm. Consider the following figure:

AAbbb  AAaaa
AAbbb  AAaaa
AAbbb  BBbbb
AAccb  BBbbb
AAccc  CCccc
DDccc  CCccc
DDcce  DDddd
DDeee  DDddd
DDeee  EEeee
DDeee  EEeee

Capital letters are Republican districts and lowercase letters are Democratic districts. Same map with two different districting plans. The five districts each have a letter.

If we define fair as a proportional result here, which is fairer? The proportion is two Republicans to three Democrats. The first is 3 to 2 Republican to Democrat. The second is 0 Republicans to 5 Democrats. The first is closer to the ideal ratio, but the second looks better to the eye.

We don't see this in the state map, as most of us don't know the geographic regions of North Carolina. So it isn't evident to us when most of these relations are broken. This creates a superficial appearance of fairness while actually creating an unfair result. Unfair to whom? Primarily the communities that get split without any recourse short of changing to a different algorithm.

Contrast this with a proportional method like Single Transferable Vote. There the voters get to choose how they group. Is ideology most important? Voters can pick ideologically similar representatives. Is local representation important? Then voters can pick representatives who live close to them (and define for themselves what close means). Minority voters can choose people with similar backgrounds. And voters can suggest how they'd like to compromise if their first choice doesn't make it.

  • 1
    Using a computer program that has no knowledge of political considerations is good. However it would be difficult to enact if the results could be predicted. I would suggest the algorithm must include at least a partial reliance on a random seed. The code would be published well in advance agreed to by both parties. Then when the census occurs a lottery type drawing would be held. The press, some guy in his basement, anyone really, would be able to immediately run the program to see the new districts. There would be no opportunity for shenanigans. – Readin Feb 12 '17 at 3:16
  • 5
    @Readin - From the linked page: My current solver takes 3-4 hours to find one solution to the California congressional map, and because every run is different due to some random factors I run the solver solid for a couple weeks to get the few best results out of it. Really, it would be better if there wasn't a random component. That way everyone could take the same inputs and produce the same results. And the question is just concerned with the mechanism, not the implementation (which you're right, would be hard). – Bobson Feb 13 '17 at 5:35
  • If you have a randomly selected seed that you publish on the appointed day, then everyone can use that same random seed as input and get the same "random" results. I think this is necessary because solution has to be a possible solution, and it isn't possible if it can't be enacted. Without requiring the solution to be possible one might just say the solution is for all politicians to put aside their partisanship and work together to draw the lines fairly. I voted for your answer alhtough I think as it stands it is an incomplete solution. – Readin Feb 14 '17 at 4:14
  • 1
    That the solver is to run multiple times and the "best" results are used is definitely a concern. The parties will not agree on which is "best" unless you have objective criteria for that (I admit I do not have time to read the whole paper. I wish I did.) – Readin Feb 14 '17 at 4:16
  • @Brythan, Re "Adding context as proposed by...": we agree that it's possible and generally undesirable to incorrectly mash-up a separate answer with an existing one. We disagree that this was in fact done in the answer which you've referenced as a model, (currently the whole question's been deleted), for your present waggish addendum. Wouldn't the general question be better discussed in Meta? – agc Nov 8 '17 at 8:12
15

One attempt that many states in the United States are trying to adopt is to take the power for creating districts out of the hands of elected political officials. Today, in most states the districts are drawn by the state assemblies and obviously they will draw the boundaries in the most advantageous way for their own party's reelection. This became even more explicitly true after the Supreme Court ruled that while it was unconstitutional to draw districts to favor candidates based on race or other protected classes, it was perfectly constitutional to use political affiliation as a deciding point.

California recently passed a ballot initiative to hand off the power for creating these districts to the California Citizens Redistricting Commission which is a nonpartisan entity that is not elected. My home state of Ohio voted in 2012 to adopt a similar policy that would see judges appoint a nonpartisan commission comprised equally of democrats, republicans and those from non-major parties. However, in Ohio the vote failed and we continue to redistrict via our elected representatives.

A different method that is more popular in Europe and at the local level in the United States is to allow for multimeter districts or at-large seats that limit the effect of the winning candidate taking the entire district and the losing candidates votes being "wasted". With multiple members being elected from the same district or votes being tallied for an at-large seat, the second place winners votes also matter, as do everyone's votes for the at-large seat. Since less votes are wasted, it becomes much more difficult to gerrymander correctly as different factions will siphon votes from each other in more unexpected ways. So another option of eliminating the problem of gerrymandering is to change the voting system itself.

  • 1
    It is known whether the California Citizens Redistricting Commission is considered 'successful'? – Mark Rogers Dec 19 '12 at 21:51
  • 2
    After it's initial passage in 2008 it was expanded to include more powers by ballot initiative in 2010. A ballot initiative to repeal it in 2010 also failed, so by the measure of popular support it has been successful. There are currently 12 states in the Union using such a method to draw districts but it depends on what you consider a "success". – Michael Kingsmill Dec 19 '12 at 22:11
  • @MarkRogers It is generally considered successful. Districts in California are a lot more competitive. – Avi Nov 14 '13 at 16:17
  • 1
    @MichaelKingsmill I was going to write my own answer but realized that this covers basically everything I wanted to say, so thumbs up, though I might describe the proportional voting system in more detail. – Avi Dec 24 '13 at 22:57
  • 4
    So long as the people on the commission are being chosen by people, there will be attempts to get the right people on the commission. See, for example, what has become of the supposedly non-partisan Supreme Court. – Readin Feb 12 '17 at 3:20
4

The following two papers suggest a solution:

The main idea is: instead of letting the governing party redistrict the entire state, let each of the two parties redistrict half of the state. Thus, gerrymandering is still possible, but the advantages of gerrymandering are divided fairly between the parties, and the total effects of gerrymandering are cancelled.

The papers explain in detail how to divide the state into two halves in a fair way, such that each party gets a fair redistricting power. In short, it is a variant of the old cake-cutting rule: "I cut, you choose".

  • 2
    Nice links. It's a good idea in theory, and it's definitely better than the majority of current systems, but it still would disenfranchise any third party(ies) which might want to get involved. – Bobson Nov 14 '13 at 14:41
  • 4
    Erel, that just reinforces incumbency. – Avi Nov 14 '13 at 16:18
  • 1
    @Bobson yes, their solution is specifically crafted for two parties. I believe this can be generalized to three or more parties, just like the "cut and choose" algorithm has been extended to any number of players. – Erel Segal-Halevi Nov 14 '13 at 18:01
  • What would you think of the idea of defining a "Gerrymander quotient" which tallies up the perimeters of district boundaries, minus some adjustments for odd-shaped boundaries that coincide with certain natural or pre-existing political features (city and county borders, for example), and then saying that anyone can submit a districting map, and the one with the lowest Gerrymander Quotient will be used? – supercat Feb 19 '14 at 0:48
  • @supercat That sounds interesting. Maybe you should write to experts in this field to get feedback. – Erel Segal-Halevi Feb 19 '14 at 6:01
3

It is possible to eliminate redistricting altogether. Redistricting is caused by geographic districts, which are inherently unfair:

  • A third to half of the voters in a district are disenfranchised because their candidate loses. So no one reflecting their beliefs represents them.

  • If you have districts that minimize boundary crossing, then the more concentrated party is disadvantaged (e.g. Democrats in urban areas).

  • If districts ignore other boundaries, then those who care about those boundaries are disadvantaged. Spread over multiple districts, they can never win representation. Or they are specially advantaged, because all the districts will have representatives pandering to them.

  • In general, special interests that only disadvantage other groups a little receive extra advantage. For example, both pro-choice and pro-life positions are minority positions in their most radical forms seen in the party platforms. This is because geographic districts require broad coalitions to govern.

  • They tend to devolve into one or two parties from which all the serious candidates come. And those that are one party districts greatly favor incumbents.

  • Politicians choose the redistricting method and therefore effectively choose their voters. Which is of course backwards.

The obvious solution to the last problem is to let voters choose the districts. Dump all the districts in a state into one pool. Let each voter choose a slate of representatives but let only one of their votes count. So each can provide a full list, but only one representative will result. There are several systems for this. The most used is currently Single Transferable Vote (STV). With STV, as the district pool grows, the number of disenfranchised voters decreases.

Take Massachusetts as an example. Because the Democrats are spread out all over the state, they win every district. They don't even have to gerrymander. It would take gerrymandering to group together enough Republican voters to make a reasonable swing district.

Blacks in South Carolina have the same problem as Republicans in Massachusetts. They dominate one urban district, but they are spread out enough throughout the state that they are part of the minority in every other district.

Libertarians have the same problem only worse. There are enough libertarians in states like California, Texas, Florida, and New York to justify a Representative, but they are spread out too much.

  • I'm not sure what you mean by "let voters choose the districts". I understand STV, but are you suggesting that everyone be elected at-large, proportionally, some other way, or do you literally mean that voters get to choose which district their vote is counted in? – Bobson Feb 11 '17 at 23:59
  • Proportionally. – Brythan Feb 12 '17 at 0:11
  • The obvious solution to the last problem is to let voters choose the districts - why not let the losing side/candidates choose? That would be an interesting turn of events... both sides would get represented (winner gets to choose the laws... loser gets to help decide the next election. And, it would go back and forth). Let the winning "nominee" of each side choose. – WernerCD Nov 6 '17 at 19:17
  • 2
    @WernerCD Because in Democracy, or rather what people outside of the US call Democracy, there are far more than 2 sides. – Peter Nov 7 '17 at 14:43
2

The state of Iowa has some standard guidelines and criteria for drawing the voting district lines.

Implementing those guidelines and criteria are done by a non-partisan board.

That's the key. Not bi-partisan, or partisan, but non-partisan.

Iowa redistricting takes partisanship out of mapmaking - The Boston Globe

2

One proposed way for the House of Representatives is to abolish districts altogether and go with Proportional Representation. Say we have the Bull Moose Party (BM), The Federalists (Fers), The Whigs (Wig), and The Prohibition Party (Pro) running in Maryland, which has 8 house seats. There would be two races in which the first one was for party representation. In this election, we have the outcomes of 50% BM, 25% Fers, 21% Wig and 4% Pro. Under most Proportional Representation, that would give 4 seats to the Bull Moose, 2 seats to the Federalists, and typically another 2 seats to the Wig (most of these systems require a 5%+ vote to take a seat, and no one loves the guys running on the No Booze platform that much, so they don't get any seats). The second election would be for the citizens to pick the number of party candidates running for those seats that they like. This could be a first past the post of all party candidates OR ranked choice (I prefer instant run-off, just cause it adds that much more power to the vote).

This second vote could be opened to all citizens to pick candidates for all parties so a Federalist can vote for a Bull Moose or a Whig that is closer to his or her own politics. It could also be done similar to the primaries where only party members are allowed to vote for the parties contribution to the delegation. At a state level its a bit more difficult as you would want someone who is ideally closer to home going to your state capitol.

This could also work for a more equitable Electoral College, where instead of a candidate winning all of the state, they only win a portion of electors that represents the out come of voting inside the state, with the remaining 2 electors going to the winner of the popular vote in the state.

At a local level, this system needs further refinement as ideally, a state representative needs to be closer to your issues than a federal level representative. Again, it could work, but I'm not sure how to resolve that issue. This system tends to favor more centralized governments, which the United States is not.

  • Perhaps you could give an example of this working in Europe, or of a proposal in the US. – user9389 Nov 6 '17 at 21:13
1

Washington State assigns the task of coming up with state and federal legislative districts to a small committee. In practice, it consists of two Democrats and two Republicans. If they fail to agree on districts, the task is assigned to the State Supreme Court. Also, the number of ways that districts can be gerrymandered is minimized by having the same districts for both the State Senate and the State House of Representatives. Each district has one Senator and two Representatives, each of whom is elected on a first-past-the-post basis. The resulting districts tend to be fairly compact; they are nowhere near as gerrymandered as King County's council districts.

1

Cross-posting from: What is the proper (non-"Gerrymandering") purpose of political districting?

First of all, let me restate your assumptions: "districts are useful" and "Gerrymandering is improper". I agree with both, but several people could argue (specially with the first one). So let us take them as axioms for this answer.

The reason why Gerrymandering is possible has to do with two facts:

  • Votes correlate very well with neighborhoods. So it is not hard to draw districts that give you a prescribed percentage of votes: for example, getting 51%, 51% and 0% would be a very good outcome for a party holding 34% of the total votes in a three-districts area.
  • The amount of freedom in drawing district lines is almost unlimited now. Parties have to draw something connected and respect a few other criteria, but that's it.

Since one cannot change the first item above, one could focus on the second. If each district had to be constructed following a limited set of rules, the problem could be solved. An example attempt would be the following:

Each district has to be delimited by either its own natural barriers or by (no more than six) straight lines

This would greatly limit how much leverage the party in power has to distort results and should solve the problem.

This proposal is very much related to the statistical concept of VC-dimension

0

There was an idea I saw that consisted of a ranked choice voting system in which House seats would go to the top candidates in a statewide election. A state is a fixed area that cannot be gerrymandered. (The ranked choice means that a vote would give partial points to a seconds and maybe third candidate. It is desirable for reasons that are not directly relevant to this question.) Of course reform in this area would be nearly politically impossible in normal times, so the chance of it happening any time soon is exactly 0.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.