SoylentNews is people
SoylentNews
The Washington Post reports Supreme Court rules race improperly dominated N.C. redistricting efforts
The Supreme Court ruled [May 22] that North Carolina's Republican-controlled legislature relied on racial gerrymandering when drawing the state's congressional districts, a decision that could make it easier to challenge other state redistricting plans.
The decision continued a trend at the court, where justices have found that racial considerations improperly tainted redistricting decisions by GOP-led legislatures in Virginia, Alabama, and North Carolina. Some cases involved congressional districts, others legislative districts.
[...] [The justices] were unanimous in rejecting one of the districts and split 5 to 3 on the other.
AlterNet reports
Republican legislators used surgical precision to pack black voters into just two districts, the tentacular 1st and the snake-like 12th. The lower court found that these districts targeted voters on the basis of race in violation of the constitution, a move that effectively prevented black voters from electing their preferred candidates in neighboring seats. map
[...] This now-invalidated congressional map was one of, if not the very most, aggressive partisan gerrymanders in modern history. North Carolina is a relatively evenly divided swing state--Donald Trump won it by just 3 points last year--yet these lines offered Republicans 10 safe districts while creating three lopsidedly Democratic seats. Amazingly, all 10 Republican districts hit a perfect sweet spot with GOP support between 55 and 60 percent, a level that is high enough to be secure yet spreads around Republican voters just carefully enough to ensure the maximum number of GOP seats possible.
(Score: 2) by NotSanguine on Thursday May 25 2017, @10:08AM (18 children)
How about we make them of equal population, given that's how districts are divided? But also create them as close to squares as possible, none of this snaking around and through based on party registrations.
No, no, you're not thinking; you're just being logical. --Niels Bohr
(Score: 5, Insightful) by SpockLogic on Thursday May 25 2017, @12:46PM (4 children)
If you want the voters to win then redistricting must be taken out of the hands of politicians. Politicians cannot be trusted not to look after their own sectarian interests above all else. An independent commission charged with creating compact, contiguous districts, that keep political units and communities within a single district would be the answer.
Overreacting is one thing, sticking your head up your ass hoping the problem goes away is another - edIII
(Score: 2) by EvilSS on Thursday May 25 2017, @01:25PM
(Score: 1, Interesting) by Anonymous Coward on Thursday May 25 2017, @01:40PM (1 child)
The alternative is what the person at the top of this thread said: eliminate districts. The constitution does not require them to exist. Each state is free to decide how to elect their representatives, which could be a state-wide at-large election along the lines of proportional representation. Or, as a half-way, in large states, make mega-districts that combine ~5 house seats into one larger district and do PR within that mega-district.
(Score: 0) by Anonymous Coward on Thursday May 25 2017, @05:54PM
It's a nice idea, but the problem is that there's generally no political will to make it happen.
If either party has even a slight majority in a given state, that party, and thus the majority of people in the state, would much rather win all the seats in Congress (or as many as jerrymaniacally possible) than, say, 6 out 11 seats. So no matter which party is on top, the majority of people will probably support jerrymandering and jerrymander-prone districting procedures. (The exception, given a two-party system, might be a state so evenly divided that nobody knows who's on top, and both sides would rather split the seats than fight for all-or-nothing. But more likely, both parties will be sure that they have the edge, and both will favor all-or-nothing.)
if you try to appeal to fairness above party interest, those who don't laugh at you will point out that all these other states (and they'll obviously pick examples that go to the opposite party) do use rules that leverage a small majority of voters into a large majority of congressional seats, and ask how it's "fair" that they unilaterally diminish their own power, giving those other states an unfair advantage.
In a functioning representative democracy, there's more than two parties, and none of them usually has a majority of voters, so this sort of math doesn't apply -- any system that could be exploited to give any one party a disproportionate share of representatives, will be opposed by the majority of people.
It's ironically depressing, then, to note that our system of winner-take-all districts instead of proportional representation is one of the main factors (though IMO a lesser one than our use of the FPP voting system for single-candidate races) that perpetuates our two-party system, and prevents us from having the functional democracy needed to muster any will to change it...
(Score: 2) by jmorris on Thursday May 25 2017, @05:34PM
What do you call someone wielding political power? The power to draw district lines is inherently political, the ones given the authority to do that will quickly become politicians even if you pick random names from a phone directory. Faster in fact if you do that since they will lack any hardening against the lobbyists who have an ax to grind in the decisions.
(Score: 3, Interesting) by Kromagv0 on Thursday May 25 2017, @01:49PM
While it wouldn't result in square districts I have thought that a simple unsupervised clustering algorithm would be the best way to do redistricting. Put a few constraints on it to try and keep cities and counties together. Initially have the sitting elected representative's home be the center of the district. Then from there let it do clustering so that each district is equal size. After that recalculate the middle of the district and rerun the clustering. Keep going until it settles down to a minimal movement. If you gain or lose seats in a stat randomly remove a congress critter or randomly drop new centers on the map and start running it. I would also suggest having a couple of runs going that use random initial district centers. Then let the state legislature vote on which one of the 3 district maps they want.
T-Shirts and bumper stickers [zazzle.com] to offend someone
(Score: 2) by AthanasiusKircher on Thursday May 25 2017, @07:18PM (2 children)
Yes, this is generally popular. But it doesn't necessarily solve gerrymandering -- it merely changes the form. Witness what happened in Indiana [wikipedia.org] after the 2010 census. If you view the historical progression of district maps there since the 1970s, it appears that Indiana mostly has moved toward more "square"(-ish) districts over time. But Democrats argued after the 2013 redistricting that it would benefit Republicans (who have, by the way, gerrymandered the state legislative districts to a high degree) and mostly cement their seats by concentrating Democratic power in few regions that happen to contain the more urban (and more liberal) areas. So it looks more "fair," but is it?
The problem, to my mind, is that we have two incompatible metrics of "fairness."
(1) One metric says we need "compact" geographic districts, ones that perhaps line up with reasonable political and/or community boundaries, so that a representative is representing a coherent group of people geographically.
(2) Another metric says we need parity between the parties, so that each party has roughly a chance of winning a number of seats equal to its respective proportion of the population STATEWIDE (or nationwide).
These two goals are diametrically opposed. When you divide up data, you can create different outcomes for subsets compared to the overall data trend. This is well-known to anyone with basic familiarity with Simpson's paradox [wikipedia.org]. So, it's quite possible to produce a mathematically geographically "perfect" division by whatever metric that is supposedly politically neutral, and STILL have a situation where 60% of the nation votes for Democrats and Republicans win 55% of House seats (or vice versa, though the fact that Democrats tend to be concentrated in high-density population zones makes it somewhat more likely that they will suffer more under most "prefer squares" metrics).
If we really want to satisfy both conditions, we'd need to do something much more radical, like allocate Congressional seats to each party on the basis of the fraction of the popular vote that each party gets nationwide (or at least statewide), as is done in some parliamentary systems [wikipedia.org]. That would eliminate much of the redistricting issue, but it would also radically change the idea that a local geographic population gets to choose its own preferred representative (and perhaps run afoul of the Constitution).
(Score: 2) by AthanasiusKircher on Thursday May 25 2017, @07:25PM
Sorry -- should have caught this before submitting, but for my first "metric" I meant: "ones that perhaps line up with reasonable municipal and/or community boundaries." (I meant "political boundaries" in the sense of county or city lines, etc., not according to political party affiliation.)
(Score: 2) by FakeBeldin on Friday May 26 2017, @01:41PM
Allow me to reformulate those:
1. Geographical districts
2. Proportional representation statewide
If you require the voting system to deliver these, then yes, these are opposed.
#2 must be handled by the voting system to preserve some amount of fairness to the results.
#1 doesn't - every candidate is free to say "I'm running on behalf of geographical region (x1,y1)-(x2,y2)! Except the folks at (x3,y3)!"
So an easy solution to the issue you raised is to have elections with proportional representation: X seats in the house, the top X of the election get those seats.
If #1 is actually important to people, then the "free market" will solve it. Candidates will find the optimum size of an area to represent and just claim that themselves. Voters who care about geographical representation might vote for them, others likely won't.
If some of those candidates are elected, then apparently there is merit to #1. If not, then either no candidate found the right way to implement #1, or voters just care about other things more than about geographical representation.
(Score: 0) by Anonymous Coward on Thursday May 25 2017, @09:50PM (8 children)
Going by area is easy. Even better, just hold a contest and let the winner be the map with it's longest perimeter being smaller than that of any other proposed map. (What is your worst-case huge district? Is yours less awful than what your competition has? If so, your map wins and you get famous.)
Going by population is hard. We can't reliably count people. We don't even know who is in the country.
Squares are not appropriate. The ideal map is mostly hexagon-like shapes. (on a curved Earth surface so not really hexagons) You vote based on the location of the center of your property.
(Score: 2) by NotSanguine on Friday May 26 2017, @03:02AM
That would be a great idea, if it weren't so moronic, given the long-held ideal of "one man, one vote" and not "one hectare, one vote." And then there's that pesky thing we call "the law" [cornell.edu] which requires 435 congressional districts split by *population* and not area. So....not so much.
Sure. hexagons would fill areas much more efficiently. My point was not to glorify any particular polygon, but point out that current congressional districts often look more like knotted up pasta than polygons. This is inappropriate in my view.
No, no, you're not thinking; you're just being logical. --Niels Bohr
(Score: 2) by NotSanguine on Friday May 26 2017, @03:02AM (6 children)
That would be a great idea, if it weren't so moronic, given the long-held ideal of "one man, one vote" and not "one hectare, one vote." And then there's that pesky thing we call "the law" [cornell.edu] which requires 435 congressional districts split by *population* and not area. So....not so much.
Sure. hexagons would fill areas much more efficiently. My point was not to glorify any particular polygon, but point out that current congressional districts often look more like knotted up pasta than polygons. This is inappropriate in my view.
No, no, you're not thinking; you're just being logical. --Niels Bohr
(Score: 2) by FakeBeldin on Friday May 26 2017, @01:52PM (5 children)
This is inherent in any form of system-enforced geographical distribution of political power (i.e., districts).
Using regular polygons is blatantly unfair: the 1000 square miles of desert with one voter gets 2 seats in the house, while the urban area with 4 towns and 1.5 million voters gets one seat.
So, in any system that is based on geographical districts, the size and shape of those districts will be tweaked. There are benign reasons for that and there are malicious reasons for that, but since voters aren't uniformly distributed and since the boundaries of the election (i.e. the state or municipality) aren't perfectly fitting to polygons, the size and shape of districts needs to be tweaked. That problem is completely avoided with proportional representation: x% of the votes gets you x% of the seats. Every. Single. Time.
(Score: 2) by NotSanguine on Friday May 26 2017, @05:30PM (2 children)
I'm not sure if you're being deliberately obtuse or if you just don't realize how redistricting/gerrymandering has worked in the U.S. [govtrack.us]
I'll give you the benefit of the doubt and assume that you're ignorant. In the U.S. congressional districts are divided *by population* and not by geography. The issues I mentioned WRT regular polygons wasn't a suggestion that such polygons be of equal area, but rather that they are of created in as compact and normalized a fashion as possible, without any consideration of *who* lives in those areas.
In too many places, congressional districts have been drawn to divide members of the party(ies) not favored by the state legislatures in order to create districts where, regardless of how they vote, they are always a minority in that district.
As such, when drawing district boundaries, intentionally splitting areas which would seem to create the most compact and normalized districts, in order to limit the voting blocs for specific parties, is malicious.
I've often thought that proportional representation in our legislative bodies was a good idea. Not because I necessarily think a parliamentary system would be superior (I think that the separation of the legislative and executive branches is an excellent idea, which is lacking in parliamentary systems), but because it would allow more diverse voices into our legislative processes and the national debate.
No, no, you're not thinking; you're just being logical. --Niels Bohr
(Score: 2) by FakeBeldin on Friday May 26 2017, @06:16PM (1 child)
Neither, actually - I was talking about the implications of districting. These hold for the USA and for any other country that uses districting.
Thanks! Unfortunately, I'm not ignorant, so sorry about that.
I was talking about theory, you're talking about practice.
In a nutshell, when districting anywhere (including in the USA):
1. population size per district does not mandate a certain structure. Districts may be rotated, moved, or squished in one direction and elongated in another.
2. Since voters are not uniformly distributed and since the borders are not nice straight lines, any zoning algorithm will make some arbitrary choices.
For an example, see this image [wp.com]. Uniform density of voters, 4 districts, every voter voting purple or green. The most obvious point of this picture is to illustrate how choosing districts carefully can determine the result. But: the key thing for this discussion is this: it is impossible to say any of these divisions is intrinsically wrong. Sure, the bottom right looks fishy - but that doesn't mean that this must never be allowed to happen anywhere. In fact, I'm sure it's possible to come up with example settings that embed this division of voters where each of the districting options depicted leads to the most fair districts in the rest of the setting.
For example, say this image depicts the city centre, and outside the city centre voter density varies. Now if you start creating your districts from the city centre, you might get districts like shown in the top row of the picture. However, in case this is the last part of the state to be districted, you could conceivably end up with the lower right division.
The UK (to name but one) must be doing it wrong then [wikipedia.org]
(Score: 2) by NotSanguine on Friday May 26 2017, @07:16PM
I suspect that we may be in violent agreement here.
My concern WRT redistricting is one of the deliberate, active weakening of one voting bloc to benefit another.
There are many ways to look at that issue. My primary point is that *preferring* regular shapes (or reasonable combinations thereof) for districts can limit (but certainly not eliminate) attempts to create district boundaries that favor one group over another.
My (albeit limited) understanding is that in most parliamentary systems (and I thought this of the UK as well, but I could be wrong), was that the majority party (or coalition) in the legislature populated the executive roles from within its own ranks and those folks retained their seats in the legislative branch (i.e., the current Prime Minister [wikipedia.org] remains the MP for Maidenhead [wikipedia.org]).
In the U.S, by contrast, Jeff Sessions [wikipedia.org] was required to resign his legislative post as a Senator from Alabama, before he could take on his role as the U.S. Attorney General.
As such, Theresa May (the head of the U.K.'s executive) will cast votes on legislation in parliament as the MP for Maidenhead. In the U.S., with the (extremely rare) exception of the Vice President casting tie-breaking votes in the Senate, no executive branch member *ever* votes on legislation.
No, no, you're not thinking; you're just being logical. --Niels Bohr
(Score: 0) by Anonymous Coward on Saturday May 27 2017, @04:46AM (1 child)
The rural area is getting screwed. Politicians find it easy to ignore the needs of the rural population. Going by area fixes it.
If you are in an urban area and want more say, you can move. Now, you might say rural people should move, but this is harmful. Cities are unable to support themselves for food and water. It all has to be trucked in. This adds pollution that usually isn't even counted toward the city causing it. (the least-polluting is smallish cities with short commutes, not urban or true rural) Having people in urban areas is a disadvantage in war; they make easy targets.
Anyway, you can't measure population well. It is estimated that California unjustly has an extra 5 seats in the house due to non-citizens being counted; this is a motivation to support crime. There is always a fight over counting the homeless.
It's no good to have people drawing the lines either. If you use county lines, people will redraw the counties.
A specified algorithm is decent, yet hard to get right. An easier solution is to make a contest out of it, specifying only a way to judge the contest entries.
(Score: 2) by FakeBeldin on Saturday May 27 2017, @02:48PM
I'm just going to focus on one part of your reply that flummoxed me:
What? That's impossible. They must know who is allowed to vote. If they know this, then they can infer the total number of eligible voters.
Now I'll grant that that's a lower bound for the total number of inhabitants of the state, but:
1. does that matter? I.e.: is the number of seats in the house dependent on total number of voters or that of inhabitants?
2. If the latter: there are surely ways to extrapolate from total number of voters to total number of inhabitants in a better way.
(2a. If the latter: why don't you abolish that and use the total number of voters instead?
Since using inhabitants should basically be scaling the number of voters, it doesn't matter so much, so you can use that unfudged
number officially, and use an official scaling factor which will apply equally to all. )