Would you support *THIS* Constitutional amendment to end gerrymandering?

[TheReckoning]

Congressional Districts are drawn by a computer algorithm to draw districts by compactness, while trying to follow county borders (for simplicity’s sake). If a party wants to challenge any given drawn district, claiming the algorithm was rigged, they can, but they have to prove it was drawn unfairly to get it changed.

[John Dule]

Yes. Democracy can be improved by the introduction of AI and algorithms, and even if people want to complain about "bias slipping in," that's still a much better system than what we have now.

[President Punxsutawney Phil]

This would be an utter trainwreck.
EDIT: I'm stupid, I didn't read the OP thoroughly. I think it'd be an improvement upon re-reading.

[President Johnson]

Yes, but I'm not sure algorithms would be difficult to write into law. The best solution remains a non-partisan commission.

[Sol]

FPTP districts should be designed to represent communities on the ground--that's the reason why single-member seats exist. Arbitrary algorithms defeat the purpose.

[John Dule]

FPTP districts should be designed to represent communities on the ground--that's the reason why single-member seats exist. Arbitrary algorithms defeat the purpose.

This sounds like gerrymandering, just with extra steps.

[Sol]

FPTP districts should be designed to represent communities on the ground--that's the reason why single-member seats exist. Arbitrary algorithms defeat the purpose.

This sounds like gerrymandering, just with extra steps.

I mean you can support dumb inhuman redistricting methods all you want, but don't complain when people dislike algorithms splitting Baltimore five ways and crossing the Cascades nonstop.

[Del Tachi]

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy. 

[John Dule]

I mean you can support dumb inhuman redistricting methods all you want, but don't complain when people dislike algorithms splitting Baltimore five ways and crossing the Cascades nonstop.

You can program algorithms to follow certain boundaries and stay compact in specific areas. This is literally a non-issue.

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy. 

What we have right now is an affront to democracy. The fact that IL-4, MD-3, TX-35, and OH-9 even exist to begin with is an affront to democracy. You can complain all you want about how "imperfect" a computer algorithm might be, but it is an obvious improvement over the status quo. The goal is not to replace an imperfect system with a perfect one, but to simply improve upon it.

[Sol]

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy.  

What we have right now is an affront to democracy. The fact that IL-4, MD-3, TX-35, and OH-9 even exist to begin with is an affront to democracy. You can complain all you want about how "imperfect" a computer algorithm might be, but it is an obvious improvement over the status quo. The goal is not to replace an imperfect system with a perfect one, but to simply improve upon it.

We literally agree on this--you can improve upon the existing system by instituting fair commissions or boundary review boards like those which exist in California or the UK!

[Oregon Eagle Politics]

Yes.

[Del Tachi]

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy. 

What we have right now is an affront to democracy. The fact that IL-4, MD-3, TX-35, and OH-9 even exist to begin with is an affront to democracy. You can complain all you want about how "imperfect" a computer algorithm might be, but it is an obvious improvement over the status quo. The goal is not to replace an imperfect system with a perfect one, but to simply improve upon it.

Those districts exist as the result of democratically-elected legislators engaging in a transparent, regular process playing out before voters and the press.  Computer-drawn districts could never meet a similar standard.

I don't question whether computer-drawn districts could perform better than legislator-drawn districts on the basis of some given "neutral" criteria.  The issue is that ceding complete control of redistricting to computers is undemocratic technocracy only for the sake of some potential marginal improvement thought up by people with think-tank brain (as you admit yourself.)

[John Dule]

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy.  

What we have right now is an affront to democracy. The fact that IL-4, MD-3, TX-35, and OH-9 even exist to begin with is an affront to democracy. You can complain all you want about how "imperfect" a computer algorithm might be, but it is an obvious improvement over the status quo. The goal is not to replace an imperfect system with a perfect one, but to simply improve upon it.

We literally agree on this--you can improve upon the existing system by instituting fair commissions or boundary review boards like those which exist in California or the UK!

The California districts are not particularly well-drawn either. The idea of nonpartisan commissions is nice in principle, but I would much prefer a more objective, easily quantifiable standard that we can apply without having to rely on a small handful of bureaucrats and their subjective opinions. 538 can draw better districts than these clowns, it seems.

[Sol]

I mean you can support dumb inhuman redistricting methods all you want, but don't complain when people dislike algorithms splitting Baltimore five ways and crossing the Cascades nonstop.

You can program algorithms to follow certain boundaries and stay compact in specific areas. This is literally a non-issue.

I'd point you to some of the experimental redistricting done by folks on the Political Geography board. Muon2 and Torie had a period of the middle part of this decade where they developed a system of redistricting using only objective defined metrics, like Jimrtex's UCCs, local government councils, county and city boundaries, etc. Not literally redistricting with an algorithm but the results they produced were similar to ones drawn by a more sophisticated computer program.

I was and still kind of am somewhat sympathetic to those efforts (and those folks are great mapmakers, don't get me wrong) but when I'd read those threads back in the day I'd get my feathers ruffled by the way in which those maps would ignore subjective but important nuances because it performed better on the metrics.

A good example of this IMO was when Muon posted a NC redistricting map which split the inner core of the city of Charlotte down the middle (I can't seem to find this map, but I will post it here once I do). It performed well on the measures which muon was using, and which seemed fine in abstraction, but it created a result which would be seen as gerrymander by anybody from Charlotte.

Ultimately I'd argue that redistricting, for all the trappings of math which tend to attach itself to it, is really more equivalent to something like graphic design.There's no objective "best map" out there, in the same way that you can't say that there's a "best novel." There are general principles out there (which would be avoiding gerrymanders/just poor maps, being democratically accountable and avoiding splitting local governments) but once you get to a certain level of quality, redistricting is basically about taste.

Lfromnj and I have talked about how in drawing Oregon, a lot seems to boil down to how exactly you decide to split Jackson County, which sits at a major choke point and is the obvious county to split if you're trying to something halfway decent. You can either put the towns around Medford or Ashland in DeFazio's district. There are reasonable arguments for both maps, both from nonpartisan perspectives (the Medford area is better connected to the west, while Ashland is more distinct from Medford) and from partisan perspectives, since how you split Jackson has a major impact on the partisan lean of DeFazio's seat. There's not really a right or wrong answer here, and the outcome you choose will simply reflect your priorities and biases.

[John Dule]

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy. 

What we have right now is an affront to democracy. The fact that IL-4, MD-3, TX-35, and OH-9 even exist to begin with is an affront to democracy. You can complain all you want about how "imperfect" a computer algorithm might be, but it is an obvious improvement over the status quo. The goal is not to replace an imperfect system with a perfect one, but to simply improve upon it.

Those districts exist as the result of democratically-elected legislators engaging in a transparent, regular process playing out before voters and the press.  Computer-drawn districts could never meet a similar standard.

I don't question whether computer-drawn districts could perform better than legislator-drawn districts on the basis of some given "neutral" criteria.  The issue is that ceding complete control of redistricting to computers is undemocratic technocracy only for the sake of some potential marginal improvement thought up by people with think-tank brain (as you admit yourself.)

You could win a gold medal in mental gymnastics if you ever went pro. Are you seriously defending the redistricting process as "transparent?" Ohio and Texas have their districts drawn by biased partisan hacks under shady circumstances. The "voters and the press" to whom you refer have repeatedly decried this process as unfair, unrepresentative, and contrary to the principles of democracy. But so long as half the electorate (aka the partisan hacks' constituents) don't care about their fellow Americans being disenfranchised, nothing will change.

In fact, an algorithm could be much more transparent than what we have now. We're not talking about something byzantine and complex like Facebook's advertisement algorithms. We're talking about a simple mathematical formula that can be used to calculate compact, equivalent districts. A second-year CS student with a Macbook Pro and Python could build that in a weekend. If you're so worried about the spoooooky computer stuff, then the algorithm could easily be published and reviewed by members of the public.

The rest of your comment is more populist whining about "technocracy," which I won't deign to address.

[Sol]

We need to shut down Silicon Valley until we can figure what's going on, folx.

I don't actually like using folx but since I'm being a douche anyway with this post I figured I oughta double down on being annoying.

[John Dule]

I'd point you to some of the experimental redistricting done by folks on the Political Geography board. Muon2 and Torie had a period of the middle part of this decade where they developed a system of redistricting using only objective defined metrics, like Jimrtex's UCCs, local government councils, county and city boundaries, etc. Not literally redistricting with an algorithm but the results they produced were similar to ones drawn by a more sophisticated computer program.

I was and still kind of am somewhat sympathetic to those efforts (and those folks are great mapmakers, don't get me wrong) but when I'd read those threads back in the day I'd get my feathers ruffled by the way in which those maps would ignore subjective but important nuances because it performed better on the metrics.

A good example of this IMO was when Muon posted a NC redistricting map which split the inner core of the city of Charlotte down the middle (I can't seem to find this map, but I will post it here once I do). It performed well on the measures which muon was using, and which seemed fine in abstraction, but it created a result which would be seen as gerrymander by anybody from Charlotte.

Dividing a city is not gerrymandering. San Francisco is broken into two districts right now; one of the border segments runs along the middle of Mission Street. Is this arbitrary? Yes. But arbitrary is not the same as unfair.

Ultimately I'd argue that redistricting, for all the trappings of math which tend to attach itself to it, is really more equivalent to something like graphic design.There's no objective "best map" out there, in the same way that you can't say that there's a "best novel." There are general principles out there (which would be avoiding gerrymanders/just poor maps, being democratically accountable and avoiding splitting local governments) but once you get to a certain level of quality, redistricting is basically about taste.

Lfromnj and I have talked about how in drawing Oregon, a lot seems to boil down to how exactly you decide to split Jackson County, which sits at a major choke point and is the obvious county to split if you're trying to something halfway decent. You can either put the towns around Medford or Ashland in DeFazio's district. There are reasonable arguments for both maps, both from nonpartisan perspectives (the Medford area is better connected to the west, while Ashland is more distinct from Medford) and from partisan perspectives, since how you split Jackson has a major impact on the partisan lean of DeFazio's seat. There's not really a right or wrong answer here, and the outcome you choose will simply reflect your priorities and biases.

I don't really disagree with any of this in principle, but the solution is not "nonpartisan" redistricting commissions, because the results they come up with are much less than ideal. Putting redistricting in the hands of a truly neutral arbiter would apply the same basic principles equally to every state, which I think is a much fairer solution.

[John Dule]

We need to shut down Silicon Valley until we can figure what's going on, folx.

I don't actually like using folx but since I'm being a douche anyway with this post I figured I oughta double down on being annoying.

Dragged kicking and screaming into the future Don't worry, everything will be fine. Daddy Bezos will handle it all from now on.

[Alcibiades]

The arguments in this thread just go to show that FPTP is a completely arbitrary and fundamentally unfair system which should be done away with. The more I see of it, the more I loathe it.

[Coolface Sock #42069]

Yes, with the only downside being that it doesn’t consider communities of common interest, but this is a black-and-white standard that can’t be manipulated, which I think is more important. Enshrining it into the Constitution means it’s settled once and for all and can’t be ignored by a president or even overturned by a Congress that doesn’t like it.

[KaiserDave]

I like people drawing the districts, specifically independent commissions.

[beaver2.0]

Yes and I would like to somehow keep communities together within counties.  For example, a county with an interstate down the middle should probably be divided along that line.

[GM Team Member and Deputy PPT WB]

as long as the code behind it is made publicly available to make it known that it is nonpartisan.

[OSR stands with Israel]

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy. 

What we have right now is an affront to democracy. The fact that IL-4, MD-3, TX-35, and OH-9 even exist to begin with is an affront to democracy. You can complain all you want about how "imperfect" a computer algorithm might be, but it is an obvious improvement over the status quo. The goal is not to replace an imperfect system with a perfect one, but to simply improve upon it.

Those districts exist as the result of democratically-elected legislators engaging in a transparent, regular process playing out before voters and the press.  Computer-drawn districts could never meet a similar standard.

I don't question whether computer-drawn districts could perform better than legislator-drawn districts on the basis of some given "neutral" criteria.  The issue is that ceding complete control of redistricting to computers is undemocratic technocracy only for the sake of some potential marginal improvement thought up by people with think-tank brain (as you admit yourself.)

The algorithm method would be chosen by the legislators when they pass the amendment

[Del Tachi]

No.  Giving control of redistricting over to anonymous data scientists and impenetrable computer algorithms preempts citizen control of the process and is an affront to democracy.  

What we have right now is an affront to democracy. The fact that IL-4, MD-3, TX-35, and OH-9 even exist to begin with is an affront to democracy. You can complain all you want about how "imperfect" a computer algorithm might be, but it is an obvious improvement over the status quo. The goal is not to replace an imperfect system with a perfect one, but to simply improve upon it.

Those districts exist as the result of democratically-elected legislators engaging in a transparent, regular process playing out before voters and the press.  Computer-drawn districts could never meet a similar standard.

I don't question whether computer-drawn districts could perform better than legislator-drawn districts on the basis of some given "neutral" criteria.  The issue is that ceding complete control of redistricting to computers is undemocratic technocracy only for the sake of some potential marginal improvement thought up by people with think-tank brain (as you admit yourself.)

You could win a gold medal in mental gymnastics if you ever went pro. Are you seriously defending the redistricting process as "transparent?" Ohio and Texas have their districts drawn by biased partisan hacks under shady circumstances. The "voters and the press" to whom you refer have repeatedly decried this process as unfair, unrepresentative, and contrary to the principles of democracy. But so long as half the electorate (aka the partisan hacks' constituents) don't care about their fellow Americans being disenfranchised, nothing will change.

In fact, an algorithm could be much more transparent than what we have now. We're not talking about something byzantine and complex like Facebook's advertisement algorithms. We're talking about a simple mathematical formula that can be used to calculate compact, equivalent districts. A second-year CS student with a Macbook Pro and Python could build that in a weekend. If you're so worried about the spoooooky computer stuff, then the algorithm could easily be published and reviewed by members of the public.

The rest of your comment is more populist whining about "technocracy," which I won't deign to address.

Exactly, the media and "voter interest" organizations decry partisan gerrymandering all the time and nothing really changes.  It must be the case then that this is an issue few voters care strongly about.  If you think there is a case to be made that transcends partisan or ideological lines, you are welcome to organize, raise money and run for office yourself on a platform of redistricting reform.  This is a feature, not a bug.

The fundamental problem with your argument here (and in most of your posts, including the horrible takes made in your own thread on trans issues lol) is a belief that technology or "science" can be truly value agnostic.  For starters, it is not a given how a computer algorithm is supposed to make tradeoffs between various "neutral" redistricting criteria (i.e., compactness, partisan proportionality, county/municipal splits, minority representation, etc.)  That's to say nothing of how subjective considerations, such as preserving communities of interest or appeasing incumbents, are supposed to be quantitatively represented in some algorithm.  You even seem to admit this issue when you say:

Dividing a city is not gerrymandering. San Francisco is broken into two districts right now; one of the border segments runs along the middle of Mission Street. Is this arbitrary? Yes. But arbitrary is not the same as unfair.

If the division of San Francisco is *truly arbitrary, then a computer algorithm has to be given some deciding criterion by which to make the chop.  That injects the subjective biases of whoever is writing the code.  Since these are inherently political questions the proper authority should be with democratically-elected representatives who are constitutionally-designed to be accessible and accountable to their constituents in a way computer codes or even non-partisan redistricting commissions cannot be.

Secondly, the relevant literature is well-established that technology adoption follows an approximate bell curve, with individuals possessing the most economic or social capital always adopting first.  Creating a redistricting algorithm may be simple for a second-year CompSci student with a MacBook, but most people are not CompSci students with MacBooks lol.  Making government processes more technical is exclusionary.  Publishing the code is of little use if most people can't understand it or legislators can't easily overrule it when it produces undesirable results.