Most important anti-gerrymandering goal

[TML]

Back in 2017-18, 538 produced its "Atlas of Redistricting" map which showed how various methods of redistricting could reshape the national House map. Specifically, they examined how districts could be gerrymandered to Democratic or Republican extremes (maximizing the number of districts where one party has a greater than 5-in-6 (~82%) chance of winning), as well as some non-partisan goals:

-Proportionality: This method tries to make districts in a state reflect the state's partisan composition proportionally. For example, if a state has 10 districts and Republicans won an average of 60% of the two-party vote share in the last two elections, the state would have six Republican-leaning districts and four Democratic-leaning districts.

-Competitiveness: This method tries to create as many "highly competitive" districts as possible - these are districts where both major parties have at least a 1-in-6 (~18%) chance of winning.

-Minority representation: This method tries to create as many "majority minority" districts as possible - districts where White voters make up less than half of the voting age population overall.

-Compactness: This method tries to create districts with "normal-looking" shapes, thereby reducing the distance between the average constituent and the district's geographic center. This was done in two different ways - one tried to respect county/municipal borders, while the other used a computer algorithm regardless of such borders. (Unlike the other three goals mentioned above, this one might not conform to the Voting Rights Act.)

Which of the aforementioned goals do you consider most important in fighting against partisan gerrymandering? If it is none of the above, please specific which other goal you consider to be most important.

[Sestak]

NOTA: "Normalness" or "Averageness".

We shouldn't be looking to establish additional conditions for map-drawers to satisfy as the solution to gerrymandering. Instead, we should focus solely on the practice itself. If you graphed the results given by all possible maps (or a representative sample of all possible maps), the most egregious gerrymanders are clear outliers - simply because they weren't decided randomly out of legal criteria, but also with the added criterion of favoring one party over the other.

The standard against gerrymandering should be a direct one - if a map is such an outlier that the only reasonable assumption is that partisan concerns were taken into account, then it should go. If it's not an outlier and falls closer to the mean, then it's fine.

[Former President tack50]

Compactness, though I can also see good arguments for competitiveness and proportionality.

[Gass3268]

Mix of Proportionality and Compactness, essentially preserving communities of interest.

[Goldwater]

Compactness, I like districts that look good and actually make some ssnse.

[Sol]

Minority Representation is the most important, but I don't think that a 50% threshold is important because it can result in packing. Instead, what's important is creating a wide range of logically-drawn districts where minorities are electable. For example, I'd prefer two black-plurality districts in Philly rather than one white district and one 50%+ black one.

[Nyvin]

Proportionality first and foremost, with a heavy secondary importance given to compactness.   If those two are done properly minority representation and competitiveness should take care of themselves.

[Strudelcutie4427]

Keeping COIs together. This will eliminate cracking and get rid of ugly lines like MD-6, MA-4, the Austinmamder, OH-9, etc

[Tintrlvr]

Keeping COIs together. This will eliminate cracking and get rid of ugly lines like MD-6, MA-4, the Austinmamder, OH-9, etc

I agree with this and voted for Compactness as a rough approximation of COIs as it's pretty unusual for a COI to be non-compact.

[cvparty]

Keeping COIs together. This will eliminate cracking and get rid of ugly lines like MD-6, MA-4, the Austinmamder, OH-9, etc

this, but actual communities of interest, not certain people’s interpretation of COIs meaning packing ppl from the other party together

[Del Tachi]

Maintaining COIs, which are always 100% subjective and often not very compact or proportional.

[Okay, maybe Mike Johnson is a competent parliamentarian.]

Proportionality first and foremost, with a heavy secondary importance given to compactness.

(believes in democracy)

[The Mikado]

Proportionality first and foremost, with a heavy secondary importance given to compactness.

(believes in democracy)

This basically breaks down with a state like yours. Proportionality would suggest a 6-3 D-R delegation is desirable for MA, but it's effectively impossible (maybe literally impossible) to draw MA with 3 GOP districts, and certainly not without an extreme gerrymander.

[Antonio the Sixth]

Other: CoIs, all the CoIs*, nothing but CoIs

"Competitiveness" is an absurd, arbitrary goal born out of mindless bipartisan fetishism that sometimes produces even worse maps than the overt partisan gerrymanders. If voters in a given community are strongly Republican or strongly Democratic, then they should be represented by a strong Republican or Democrat. Throwing an equal number of hyperpartisans from either side into a district to make it "competitive" doesn't reduce polarization: on the contrary, it only increases the tendency of elections to revolve around base turnout/voter suppression rather than persuasion.

Proportionality is just not what district elections are for. They are fundamentally incapable of producing it, no matter how hard you try. Even if you manage to achieve a situation when at a given election, a 60/40 PV split produces a 60/40 seat split, next election when the PV changes to 55/45 or 65/35, the seats are never going to exactly follow suit. If you want proportionality, there's only one solution: switch to PR (which I strongly support doing, of course),

Minority representation is a noble goal in theory, and that's one thing a CoI-based approach takes into account, but it can't and shouldn't be taken to the absurd extremes that VRA districts from the 90s and 00s took it. There are other CoIs whose needs need to be balanced with those of ethnic minorities, such as partisan communities (a typical issue with extreme VRA districts is that they actually make life worse for the party ethnic minorities actually support) or geographic areas (the typical example of the VRA district linking black neighborhoods from two different cities).

Compactness is an important guiding force, not because aesthetically pleasing districts are a goal in itself (although I'll freely admit I'm biased in their favor) but because it's usually an indication that the map is being drawn without a partisan agenda. But obviously there are times when districts need to be made less compact for the sake of better representing constituencies that aren't spread out in a nice-looking square or circle. Also, those algorithmic district maps that sometimes get bandied around are abject monstrosities that make a mockery of the redistricting process, usually slicing cities into pieces like a mosaic.


*Well, obviously it's impossible to design a map that optimally represents ALL CoIs, but there should be a goal of doing as good a job as possible.

[lfromnj]

Compactness>Competetive>>Proportionality>>>>Minority districts

Minority districts are absurd unless they are compact.

[Sestak]

Proportionality first and foremost, with a heavy secondary importance given to compactness.

(believes in democracy)

This basically breaks down with a state like yours. Proportionality would suggest a 6-3 D-R delegation is desirable for MA, but it's effectively impossible (maybe literally impossible) to draw MA with 3 GOP districts, and certainly not without an extreme gerrymander.

It’s literally impossible to draw even one R district in MA, the votes are too spread out.

[Nyvin]

Wow this thread is borked now, wtf.

[beesley]

I'd argue for a combination of geography and minority representation - communities of interest.